Events Archive before Oct 2011

  • Posted on: 17 May 2014
  • By: hadmin

Student Seminar on Tileable Bidirectional Texture Function for 3D Objects

Student Seminar

Title: Tileable Bidirectional Texture Function for 3D Objects
Speaker: Mr. PANG Wai Man
PhD Candidate
Dept. of Computer Science and Engineering
The Chinese University of Hong Kong
Date: 3 April 2008 (Thursday)
Time: 4:30 pm - 5:30 pm
Venue: ERB513

Abstract:
Conventional approach of applying bidirectional texture function (BTF) on surface of 3D objects tries to synthesize BTF directly on the surface of geometry. Applying another BTF dataset is therefore required to repeat the whole synthesis process again which is time-consuming. In this seminar, a novel systematic approach to dress BTF on 3D object surface is introduced. First, tiles of BTF are automatically created using an improved texture synthesis technique suitable for high dimensional BTF data. It consists of four major sub-steps: corner sampling, edge synthesis, frame construction, and interior area synthesis. To facilitate the rendering on GPU, BTF tiles are compressed using doubly projected spherical harmonics to make it compact enough to store on GPU memory. The synthesized
BTF tile sets can be instantly applied on the geometry seamlessly. Our experiments demonstrate our approach works well in decoupling appearance from geometry and applicable for various surface geometries. Several results of illumination under various lighting conditions, e.g. local and environment lighting, will be demonstrated during the seminar.

Biography:
Wai-Man PANG is a PhD candidate in the Department of Computer Science & Engineering, Chinese University of Hong Kong supervised by Professor Pheng-Ann HENG and Professor Tien-Tsin WONG. He has visited the Internet Graphics Group in Microsoft Research Asia Beijing from July to September 2007. His current research interests focus on non-photorealistic rendering, imagebased relighting, and GPU/PPU hardware accelerated methods.

 

Light refreshment is provided by 4:00 pm.

Please download the poster here.

 


 

Student Seminar on Speech Recognition with Multi-space Distribution Prosody Model: a Comparative Study between Spontaneous and Read Speech

Student Seminar

Title: Speech Recognition with Multi-space Distribution Prosody Model: a Comparative Study between Spontaneous and Read Speech
Speaker: YEUNG Yu Ting
  Master Student
  Dept. of Electronic Engineering
  The Chinese University of Hong Kong
Date: March 6, 2008 (Thursday)
Time: 4:30 pm
Venue: Room 513, William M.W. Mong Engineering Building, The Chinese University of Hong Kong

Abstract:
Automatic speech recognition is challenging for Chinese languages like Mandarin because of their tonal nature. A comparative study of prosodic features between Mandarin spontaneous speech and read speech is carried out in the context of automatic speech recognition. The analysis is based on a unique speech corpus that contains similar amounts of read and spontaneous speech data from the same group of speakers. Tone contours, duration of syllable and sub-syllable units are analyzed by statistical
methods. Different approaches to incorporating prosodic features into acoustic modeling are investigated in large-vocabulary speech recognition experiments. One of the key problems being addressed is the modeling of F0 for unvoiced speech. We use the technique of multi-space distribution (MSD) to handle explicit voicing status labels. For spontaneous speech, the tonal-syllable error rate is reduced from the MFCC baseline of 64.8% to 59.4% after using prosodic features with the MSD approach. For read speech, the error rate decreases from 46.0% to 36.4%.

About the speaker:
YEUNG Yu Ting received his BEng degree (first-class honor) in Electronic Engineering from the Chinese University of Hong Kong in 2006. He is currently a year 2 Master of Philosophy student from the DSP and Speech Technology Laboratory, Department of Electronic Engineering. During June – August 2007, he was a visiting student in the Speech Research Group, Microsoft Research Asia in Beijing. His research interest is large-vocabulary continuous speech recognition for conversational Cantonese and Cantonese-English code-mixing speech.

Inquiries: Linda MA (tsma@se.cuhk.edu.hk), +852-26098304 Details: http://www.cuhk.edu.hk/ms-cu-jl

Please download the poster here.

 


 

Student Seminar on Image Compression by Inverse and Forward Texture Syntheses

Student Seminar

Title: Image Compression by Inverse and Forward Texture Syntheses
Speaker: Mr. Zhen Yu WEI
  PhD Candidate
  Dept. of Electronic Engineering
  The Chinese University of Hong Kong
Date: February 21, 2008 (Thursday)
Time: 4:30 pm
Venue: Room 513, William M.W. Mong Engineering Building, The Chinese University of Hong Kong

Abstract:
In this presentation, an image compression scheme is proposed towards the perceptual visual quality instead of the pixel-wise fidelity, which is based on the recent advances in texture synthesis. In our scheme, a given natural texture image is first summarized as a small size sample by utilizing the inverse texture synthesis technique. This sample contains enough detailed information of the original texture for reconstruction. Then the assistant parameters mapping from the original texture to the sample are extracted by using a motion-estimation-like technique at the encoder side. At the decoder side, the image is reproduced by forward texture synthesis, relying on both the compressed sample and assistant parameters. Experimental results show that our proposed compression scheme can achieve  up to 56% bit-saving compared with JPEG at the similar perceptual visual quality level. And at the same bit-rate, the perceptual quality of our method outperforms those of JPEG and JPEG2000.

About the speaker:
WEI Zhen Yu is now a year 3 PhD candidate from the Department of Electronic Engineering, the Chinese University of Hong Kong. He received the B.S. and M.S. degrees, both in Electronic Engineering from Tsinghua University, Beijing, China, in 2002 and 2005, respectively. From June to August, 2007, he was a visiting student at Microsoft Research Asia (MSRA), Beijing, China. Currently he is a member of the Visual Signal Processing and Communication Laboratory (VSPC) and works under the supervision of Professor King N. Ngan. His research interests include image and video coding, perceptual image and video processing, embedded system application.

ALL ARE WELCOME

* Light refreshment will be served after 4:00p.m.

Please download the poster here.

Inquiries: Linda MA (tsma@se.cuhk.edu.hk), +852-26098304 Details: http://www.cuhk.edu.hk/ms-cu-jl
 

 


 

Student Seminar on Unsupervised Method to Incorporate Semantic Meanings with an Anchor Space for Improving Story Segmentation

Student Seminar

Title: Unsupervised Method to Incorporate Semantic Meanings with an Anchor Space for Improving Story Segmentation
Speaker: Mr. Kelvin CHAN
  Master Student
  Dept. of Systems Engineering and Engineering Management 
  The Chinese University of Hong Kong
Date:     January 24, 2008 (Thursday)
Time:     4:30p.m.
Venue:  Room 513, William M.W. Mong Engineering Building,
              The Chinese University of Hong Kong

Abstract:
News story segmentation is the task of segmenting a stream of multimedia news video/audio into news stories. It is an important prerequisite of other retrieval related tasks in the domain of news information retrieval. This seminar introduces the work that uses a semantic anchor space to improve segmentation of text transcripts of MSNBC news. This work highlights the importance of representing the underlying text in a certain form of semantic representation. We apply clustering on auxiliary source of news data to form the vectors of an anchor space, onto which the original word vectors in the news transcript are mapped. A modified version of the traditional algorithm for text segmentation is then applied, and the results are compared with those by traditional algorithms. Our work exhibits improvement of segmentation performance under some conditions, and it sheds light on directions for possible future improvements of segmentation tasks.

About the speaker:
CHAN Shing Kai is now a year 2 master of philosophy student from the Department of Systems Engineering and Engineering Management, the Chinese University of Hong Kong. He was graduated from the same university with first class honor in Computer Engineering in 2006.  Currently a member of the Human-Computer Communications Laboratory (HCCL), he works under the supervision of Professor Helen MENG on the topic of automatic story segmentation of Chinese broadcast news. His paper entitled “Modeling the Statistical Behavior of Lexical Chains to Capture Work Cohesiveness for Automatic Story Segmentation” was published in the INTERSPEECH 2007 held in Antwerp, Belgium. His research interests include information retrieval, natural language processing and speech processing.

ALL ARE WELCOME
* Light refreshment will be served after 4:00p.m.

Please download the poster here.

 


 MSRA Summer Internship Opportunities (2008)

It is time for us to nominate our students to join this year's summer internship at Microsoft Research Asia. We plan to nominate up to 2 candidates for consideration by MSRA. For nominations from joint labs, they will conduct phone interviews in order to enable them to liaise with the various group managers to get their internship slots for proper placement of interns.
To initiate the process, please send your nominations (before Tuesday, April 22, 2008) to your department representative in the MS-CU-JL management committee for your convenience, they are:

  • Professor Ronald Chung (MAE) 
  • Professor Pheng Ann Heng (CSE) 
  • Professor Lee Tan (EE)
  • Professor Dah-Ming Chu (IE) 
  • Professor Wai Lam (SEEM)

Here's more information about the MSRA internships:  

  1. Nominated candidates should be PhD students with good research performance. Masters students who have very strong research performance will also be considered.
  2. The Microsoft Fellowship awardee is entitled to an internship opportunity.
  3. Candidates should have research backgrounds that fall within the research themes of the MS-CU-JL (please refer to http://www.cuhk.edu.hk/ms-cu-jl). If the nomination includes a suggestion of an MSRA researcher who may provide mentorship, that may be helpful for consideration on the MSRA side.
  4. Applications / nominations should include (soft copies): the student's CV, transcript, published papers, research theme, a brief research statement.
  5. MSRA summer internships typically run for 10-12 weeks during the summer. The internship period of thie year will be from May to July in order to avoid The Olympic Games in Beijing.
  6. MSRA covers for the round-trip airfare between Hong Kong and Beijing, local accommodation, meals and a monthly stipend.

Please note that students should apply in-residence leave individually for the internship period. A checklist with application form is available here.

 


 

Two Section Short Course

Two Section Short Course on
1) Is there anything left for advanced video coding?
2) Emerging visual capturing and display technologies: challenges and future trends

by Professor C.-C. Jay KUO
University of Southern California
Photo Album
Slides on Flexible Resolution Conversion (set 1)(set 2)
Slides on Advanced Video Coding (set 1)

 


 

Summer Course by Professor Alon Orlitsky

Summer Course by Professor Alon Orlitsky
Department of Electrical and
Computer Engineering and Department of Computer Science and Engineering
University of California, San Diego

The title is "Universal Compression and Probability Estimation"
Photo Album

 


 

Seminar on Convergence on Communication and Broadcasting Systems

Seminar on Convergence on Communication and Broadcasting Systems
by Professor Hong-goo Kang, Yonsei University, Korea

 


 

Seminar on A Model for Conversational Robot Intelligence based on Multiple Experts and Its Applications

Seminar on A Model for Conversational Robot Intelligence based on Multiple Experts and Its Applications
by Dr. Mikio Nakano
Senior Researcher
Honda Research Institute Japan Co., Ltd.

 


 

Seminar on Multimodal Human-Machine Communication in Smart Environments

Seminar by Professor Gerhard Rigoll
Institute for Human-Machine Communication
Munich University of Technology
Germany

The title is "Multimodal Human-Machine Communication in Smart Environments"

 


 

Seminar on LiquidSilver: Towards Seamless Media Ecosystem (Introduction of Media Research at MSRA)

Dr. Shipeng LI from Microsoft Research (Asia) visited the Chinese University of Hong Kong.
Dr. LI gave a talk on "LiquidSilver: Towards Seamless Media Ecosystem (Introduction of Media Research at MSRA)"

 


 

Seminar on Image-Based Rendering of Dynamic Scenes

Dr. S. B. KANG from Microsoft Research (Redmond) visited the Chinese University of Hong Kong.
Dr. Kang gave a talk on "Image-Based Rendering of Dynamic Scenes"

 


 

MSRA Summer Internship Opportunities (2007) 4/20/2007 12:00:00 AM 5/4/2007 12:00:00 AM

It is time for us to nominate our students to join this year's summer internship at Microsoft Research Asia. We plan to nominate up to 6 candidates for consideration by MSRA. For nominations from joint labs, they will conduct phone interviews in order to enable them to liaise with the various group managers to get their internship slots for proper placement of interns.
To initiate the process, please send your nominations (before May 4, 2007) to your department representative in the MS-CU-JL management committee -- for your convenience, they are: Professor Ronald Chung (MAE), Professor Pheng Ann Heng (CSE), Professor Lee Tan (EE), Professor Dah-Ming Chu (IE), Professor Wai Lam (SEEM). Here's more information about the MSRA internships:

  1. Nominated candidates should be PhD students with good research performance. Masters students who have very strong research performance will also be considered.
  2. The Microsoft Fellowship awardee is entitled to an internship opportunity.
  3. Candidates should have research backgrounds that fall within the research themes of the MS-CU-JL (please refer to www.cuhk.edu.hk/ms-cu-jl). If the nomination includes a suggestion of an MSRA researcher who may provide mentorship, that may be helpful for consideration on the MSRA side.
  4. Applications / nominations should include: the student's CV, transcript, published papers, research theme, a brief research statement.
  5. MSRA summer internships typically run between June to August. MSRA covers for the round-trip airfare between Shenzhen and Beijing, local accommodation, meals and a monthly stipend. The joint lab will apply for in-residence leave for our students for the internship period.

A checklist with application form is available here.

 


 

MSRA Summer Internship Opportunities (2006) 3/23/2006 12:00:00 AM 4/7/2006 12:00:00 AM

It is time for us to nominate our students for the five summer internship slots at Microsoft Research Asia. To initiate the process, please send your nominations (before April 7, 2006) to your department representative in the MS-CU-JL management committee -- for your convenience, they are: Professor Ronald Chung (ACAE), Professor Pheng Ann Heng (CSE), Professor King Ngan (EE), Professor Dah-Ming Chu (IE), Professor Wai Lam (SEEM). Here's more information about the MSRA internships:

  1. Nominated candidates should be PhD students with good research performance. Masters students who have very strong research performance will also be considered.
  2. The Microsoft Fellowship awardee is entitled to an internship opportunity.
  3. Candidates should have research backgrounds that fall within the research themes of the MS-CU-JL (please refer to www.cuhk.edu.hk/ms-cu-jl). If the nomination includes a suggestion of an MSRA researcher who may provide mentorship, that may be helpful for consideration on the MSRA side.
  4. Applications / nominations should include: the student's CV, transcript, published papers, research theme, a brief research statement.
  5. MSRA summer internships typically run between June to August. MSRA covers for the round-trip airfare between Hong Kong and Beijing, local accommodation, meals and a monthly stipend. The joint lab will apply for in-residence leave for our students for the internship period.

A checklist with application form is available here.

NULL 0 1 50 1 MEETING 4/30/2008 6:14:02 PM 4/30/2008 6:14:02 PM W3C Voice Browser Working Group Face-to-face Meeting 7/11/2006 12:00:00 AM 7/13/2006 12:00:00 AM

W3C Voice Browser Working Group (VBWG) Internationalization of Speech Synthesis Markup Language (SSML) 1.1 Face-to-Face Meeting
Held at Shun Hing Institute of Advanced Engineering
The Chinese University of Hong Kong

 


 

The Second Beijing-Hong Kong International Doctoral Forum 7/3/2007 12:00:00 AM 7/6/2007 12:00:00 AM

The Second Beijing-Hong Kong International Doctoral Forum 2007 on Network and Media Computing
Held at The Chinese University of Hong Kong, Hong Kong SAR, China
Please visit http://wiki.cse.cuhk.edu.hk/PHDForum2007 for more information.

 


 

The 1st Beijing-Hong Kong Doctoral Forum 7/3/2006 12:00:00 AM 7/3/2006 12:00:00 AM

The 1st Beijing-Hong Kong Doctoral Forum 2006 on Network and Media Computing
Held at Tsinghua University, Beijing, China
Call for Papers
Forum Schedule
Photo and Video Albums
Article in the University Marketplace

 


 

 Student Seminar on Model-based Speech Separation with Single-microphone Input 5/9/2007 12:00:00 AM 5/9/2007 12:00:00 AM

Student Seminar on Model-based Speech Separation with Single-microphone Input
by Miss Yvonne LEE, Department of Electronic Engineering

 


 

 Student Seminar on Suggesting Relevant Cross-Lingual Queries by Mining Query Log of Different Languages 4/4/2007 12:00:00 AM 4/4/2007 12:00:00 AM

Student Seminar on Suggesting Relevant Cross-Lingual Queries by Mining Query Log of Different Languages
by Mr. Wei GAO, Department of Systems Engineering and Engineering Management

 


 

Student Seminar on 3D Object-based Scalable Wavelet Video Coding with Boundary Effect Suppression 3/14/2007 12:00:00 AM 3/14/2007 12:00:00 AM

Student Seminar on 3D Object-based Scalable Wavelet Video Coding with Boundary Effect Suppression
by Mr. Yu LIU, Department of Electronic Engineering

 


 

Student Seminar on A Multi-Pass LVCSR Framework: Decode Speech through Initial Recognition, Error Detection, and Error Correction 2/28/2007 12:00:00 AM 2/28/2007 12:00:00 AM

Student Seminar on A Multi-Pass LVCSR Framework: Decode Speech through Initial Recognition, Error Detection, and Error Correction
by Ms. Zhengyu ZHOU, Department of Systems Engineering and Engineering Management

 


 

Student Seminar on Ranking on Multiple Class Data 1/24/2007 12:00:00 AM 1/24/2007 12:00:00 AM

Student Seminar on Ranking on Multiple Class Data
by Mr. Wei LIU, Department of Information Engineering

 


 

Student Seminar on The Joint Interpretation of Speech and Pen-based Gestures in Beijing Transportation Information Domain 3/29/2006 12:00:00 AM 3/29/2006 12:00:00 AM

Student Seminar Series - VI: HUI Pui-Yu , Ph.D. Candidate
Department of Systems Engineering and Engineering Management.

Miss HUI gave a talk on "The Joint Interpretation of Speech and Pen-based Gestures in Beijing Transportation Information Domain"

 


 

Student Seminar on Learning Distance Functions for Multimedia Retrieval and Web Search 3/23/2006 12:00:00 AM 3/23/2006 12:00:00 AM

Student Seminar Series - V: HOI Chu-Hong, Ph.D. Candidate
Department of Computer Science and Engineering.

Mr HOI gave a talk on "Learning Distance Functions for Multimedia Retrieval and Web Search"

 


 

Student Seminar on Separating Mixed Speech Sources with Single-microphone Input using Human Perceptual Cues 3/9/2006 12:00:00 AM 3/9/2006 12:00:00 AM

Student Seminar Series - IV: LEE Siu-Wa, Ph.D. Candidate
Department of Electronic Engineering.

Miss LEE gave a talk on "Separating Mixed Speech Sources with Single-microphone Input using Human Perceptual Cues"

 


 

Student Seminar on Audio-guided Video Based Face Recognition 2/24/2006 12:00:00 AM 2/24/2006 12:00:00 AM

Student Seminar: LI Zhifeng, Ph.D. Candidate
Department of Information Engineering.

Mr LI gave a talk on "Audio-guided Video Based Face Recognition"

 


 

Student Seminar on Leaky Key Picture Reference for JSVC Transmission with Enhanced Per-pixel Error Concealment 2/5/2006 12:00:00 AM 2/5/2006 12:00:00 AM

Student Seminar Series - III: LI Jie, Ph.D. Candidate
Department of Electronic Engineering.

Mr LI gave a talk on "Leaky Key Picture Reference for JSVC Transmission with Enhanced Per-pixel Error Concealment"

 


 

Student Seminar on A Word-lattice Indexing Method for Spoken Document Retrieval & a Comparative Study of Discriminative Methods for Reranking LVCSR N-best Hypotheses 1/19/2006 12:00:00 AM 1/19/2006 12:00:00 AM

Student Seminar Series - II: ZHOU Zhengyu, Ph.D. Candidate
Department of Systems Engineering and Engineering Management.

Ms ZHOU gave a talk on "A Word-lattice Indexing Method for Spoken Document Retrieval & a Comparative Study of Discriminative Methods for Reranking LVCSR N-best Hypotheses"

 


 

Student Seminar on Learning Users' Interest to Assist Image Browsing and Searching 9/5/2005 12:00:00 AM 9/5/2005 12:00:00 AM

Student Seminar Series - I: LIU Hao, Ph.D. Candidate
Department of Information Engineering.

Mr LIU gave a talk on "Learning Users' Interest to Assist Image Browsing and Searching"

 


 

Distinguished Lecture on The Future of Spoken Language Processing: Where Do We Go From Here? 11/22/2007 12:00:00 AM 11/22/2007 12:00:00 AM

Distinguished Lecture on "The Future of Spoken Language Processing: Where Do We Go From Here?"
by Professor Roger Moore
Chair of Spoken Language Processing
Speech and Hearing Research Group (SPandH)
Department of Computer Science, University of Sheffield

 


 

Distinguished Lecture on The Multimedia Communications Revolution of the 21st Century 6/28/2007 12:00:00 AM 6/28/2007 12:00:00 AM

Distinguished Lecture on "The Multimedia Communications Revolution of the 21st Century
by Professor Lawrence Rabiner
Center for Advanced Information Processing (CAIP)
Rutgers University, USA
Presentation PowerPoint

 


 

 Distinguished Lecture on Pattern Recognition: Bayes or not Bayes, Is that the Question? 12/2/2006 12:00:00 AM 12/2/2006 12:00:00 AM

Distinguished Lecture on "Pattern Recognition: Bayes or not Bayes, Is that the Question?"
by Professor Fred Juang,
Motorola Foundation Chair Professor and
Georgia Research Alliance Eminent Scholar
Digital Signal Processing Group and Telecommunications Group
School of Electrical and Computer Engineering
Georgia Institute of Technology, USA
Photo Album

 


 

Shun Hing Distinguished Lecture on Research 2.0 11/28/2006 12:00:00 AM 11/28/2006 12:00:00 AM

Shun Hing Distinguished Lecture on "Research 2.0"
by Professor Harry SHUM,
Managing Director of Microsoft Research Asia
Photo Album

 


 

Invited Lecture on Multilingual Spoken Dialog Systems 9/28/2006 12:00:00 AM 9/28/2006 12:00:00 AM

Invited Lecture at the Harbin Institute of Technology Shenzhen Graduate School, MSRA 2006 Summer Course on Next Generation Internet Information Processing Technologies

on the topic of "Multilingual Spoken Dialog Systems"
by Professor Helen MENG

NULL 0 1 70 1 LECTURE 4/30/2008 6:43:07 PM 4/30/2008 6:43:07 PM Distinguished Lecture on Generalized Principal Component Analysis with applications to Vision and Image Compression 2/22/2006 12:00:00 AM 2/22/2006 12:00:00 AM

Distinguished Lecture by Professor Shankar SASTRY
Director
Center for Information Technology Research in the Interest of Society
NEC Distinguished Professor of Engineering
Departments of Electrical Engineering and Computer Sciences and Bioengineering
University of California
Berkeley
Professor SASTRY will give a lecture on "Generalized Principal Component Analysis with applications to Vision and Image Compression"
Guide to Lecture Venue[CUHK Campus Map]
Photo Album

 


 

The First Microsoft Joint Laboratory Symposium 12/1/2006 12:00:00 AM 12/1/2006 12:00:00 AM

The First Microsoft Joint Laboratory Symposium
Program Rundown
Photo Album

 


 

MSRA Faculty Summit 2005 in Hangzhou 10/30/2005 12:00:00 AM 10/31/2005 12:00:00 AM

Faculty from The Chinese University of Hong Kong attended MSRA Faculty Summit 2005 - Hangzhou

Summit website

 


 

Mini Exhibition - A brief introduction to research projects in the Faculty of Engineering, CUHK 11/4/2005 12:00:00 AM 11/4/2005 12:00:00 AM

Mini Exhibition to visitors from Microsoft - A brief introduction to research projects in the Faculty of Engineering, The Chinese University of Hong Kong
booklet [2.6M]
Photo Album

Dean of Engineering, CUHK (Prof. Peter Yum) making a welcome speech to the guests.Mr. Chen Zhenzhong receives the Microsoft Fellowship.
 

Multi-modal user authentication system.Network coding.Video coding.Virtual human anatomy and acupuncture.Single wheel robot.Server-less video streaming.Robotic access point for wireless network.Precomputed lighting for real-time computer games.Augmented reality computing arena for digital entertainment.Text mining from online multilingual news.

 


 

Inauguration Ceremony of the Microsoft-CUHK Joint Laboratory for Human-Centric Computing and Interface Technologies 5/4/2005 12:00:00 AM 5/4/2005 12:00:00 AM

Inauguration Ceremony of the Microsoft-CUHK Joint Laboratory for Human-Centric Computing and Interface Technologies
[Press release]
[香港中文大學與微軟亞洲研究院成立聯合實驗室]

ceremony
From left to right: Managing Directors of the joint laboratory (Professor Peter Yum, Dr. Hsiao-Wuen Hon), Vice-Chancellor of the Chinese University of Hong Kong (Professor Lawrence Lau), Chairman of International Advisory Board (Professor Andrew Yao).

honourary guests
Honourary guests at the ceremony.

 


 

Visit to Microsoft Research, Redmond 9/13/2005 12:00:00 AM 9/13/2005 12:00:00 AM

CUHK Delegation visited Microsoft Research in Redmond

 


 

Sparsity Penalized Volumetric Imaging 6/25/2008 12:00:00 AM 6/25/2008 12:00:00 AM

Faculty of Engineering & CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies

Distinguished Lecture Series 2008 by Professor Alfred Hero

We are delighted to have invited Professor Alfred Hero, Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor to visit The Chinese University of Hong Kong and deliver a distinguished lecture. Professor Hero was the President of the IEEE Signal Processing Society (2006-2008). He will share with us his research findings and his talk is entitled "Sparsity Penalized Volumetric Imaging". You are cordially invited to this lecture. The details are as follows:
DATE: June 25, 2008 (Wednesday)
TIME: 4:00p.m. - 5:00p.m.
VENUE: LT4, Lady Shaw Building, CUHK
SPEAKER: Professor Alfred Hero
TITLE OF THE TALK: Sparsity Penalized Volumetric Imaging
Abstracts:
Three dimensional tomographic imaging is an important imaging application that naturally involves sparsity constraints. For example, see-through-wall radar imaging and magnetic resonance force microscopy can be interpreted as object domain sparsity penalized inverse problems. When the forward operator is only partially known, the blind sparsity constrained inverse problem becomes relevant. We will describe approaches to solving these naturally sparse inverse problems that are based on statistical object and measurement modeling, sparsity penalization, and optimization.
Biography of Speaker:
Alfred O. Hero III received the B.S. (summa cum laude) from Boston University (1980) and the Ph.D from Princeton University (1984), both in Electrical Engineering. Since 1984 he has been with the University of Michigan, Ann Arbor, where he is a Professor in the Department of Electrical Engineering and Computer Science and, by courtesy, in the Department of Biomedical Engineering and the Department of Statistics. His recent research interests have been in areas including: inference in sensor networks, adaptive sensing, bioinformatics, inverse problems, and statistical signal and image processing. He is a Fellow of the Institute of Electrical and Electronics Engineers (IEEE) and has received a IEEE Signal Processing Society Meritorious Service Award (1998), a IEEE Signal Processing Society Best Paper Award (1998), and the IEEE Third Millennium Medal (2000). Alfred Hero was President of the IEEE Signal Processing Society (2006-2008).

***** ALL ARE WELCOME *****
* light refreshment will be available after 3:30pm. *
Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), +852-26098304

Poster
 

 


 

Microsoft Research Asia 10th Anniversary Innovation Forum 8/12/2008 12:00:00 AM 8/12/2008 12:00:00 AM

Microsoft Research Asia 10th Anniversary Innovation Forum

8:00pm -9:30pm, Tuesday, August 12, 2008
Convention Hall 2/F, Hong Kong Convention and Exhibition Centre
 
The "Microsoft Research Asia 10th Anniversary Innovation Forum" will be held in Hong Kong on August 12, 2008. This Forum is to celebrate the 10th Anniversary of Microsoft Research Asia (MSRA) and its efforts on "turning ideas into reality" for the past decade. More importantly, it brings together some of the best visionaries to explore new avenues to make a contribution to our world. At the Forum, the Chairman of Microsoft Corporation, Dr. Bill Gates, will deliver a keynote speech.

Attendance to this Forum is by invitation only and The Chinese University of Hong Kong (CUHK) is honored to have been invited by MSRA. Details of the Forum are as follows:

DATE:    12 August, 2008 (Tuesday)

TIME:     8:00 p.m. - 9:30 p.m.
VENUE: Convention Hall, 2/F, HK Convention & Exhibition Centre, 1 Expo Drive, Wanchai, Hong Kong SAR

        Program Rundown:
Keynote Speech:
Dr. Bill Gates, Chairman, Microsoft Corporation

Panel Discussion:
Panelists:
Dr. Bill Gates, Chairman, Microsoft Corporation
Professor Paul Ching-Wu Chu, President, The Hong Kong University of Science and Technology
Professor Lawrence J. Lau, Vice-Chancellor and President, The Chinese University of Hong Kong
Professor Lap-Chee Tsui, Vice-Chancellor and President, The University of Hong Kong
Moderator:
Dr. Hsiao-Wuen Hon, Managing Director, Microsoft Research Asia


Please note that there are 250 free tickets reserved for staff and students of CUHK. Due to the limited number of seats, enrollment will be on a first-come, first-served basis. For more details, interested members of CUHK may visit the event web page (http://www.microsoft.com/hk/mscorp/msra10thanniversary/default.aspx). Online registration will be available between July 2-6, 2008 on the event web page.
Important Dates:
Online Registration
July 2 - 6, 2008
Email Confirmation by Microsoft
around July 10, 2008
Online Submission of Questions to Bill Gates
from July 15, 2008
Ticket Distribution
After Mid-July, 2008
10th Anniversary Innovation Forum
August 12, 2008
 

***** First-Come, First-Served *****

7/1/2008 12:00:00 AM 0 1 84 1 OTHER 9/11/2008 6:13:36 PM 9/18/2008 11:36:02 AM Mini Exhibition - A brief introduction to research projects in the Faculty of Engineering, CUHK 

 


 

The 3rd Beijing-Hong Kong Doctoral Forum 2008 12/21/2008 12:00:00 AM 12/24/2008 12:00:00 AM

 

 The 3rd Beijing-Hong Kong Doctoral Forum 2008

on
 Network and Media Computing
December 21-24,2008
Held at Tsinghua University, Beijing, China

Call for Papers - Extended Paper Submission Deadline to

October 25, 2008

Event Highlights:
-    Paper presentations
-    Invited talks
-    Visiting hi-tech companies and research institutes
-    Local tours
-    Social activities

 Important Dates
    Paper submission deadline           Oct.25, 2008
    Notification of acceptance             Nov.1, 2008   
    Camera-ready paper due date      Nov.15, 2008   
    Forum                                            Dec.21, 2008

Please visit http://phd2008.cs.tsinghua.edu.cn for more information.

 

 

 


Seminar on Context Dependent LM Interpolation and Adaptation 9/30/2008 12:00:00 AM 9/30/2008 12:00:00 AM

**Seminar**
CUHK MoE-Microsoft Key Laboratory of

        Human-Centric Computing and Interface Technologies

------------------------------------------------------------------------------------------------

 

Title:                            

Context Dependent LM Interpolation and Adaptation

Speaker:

Dr. Andrew LIU
Postdoctoral Research Associate of the Machine Intelligence Laboratory
Department of Engineering
Cambridge University

Date:

September 30, 2008 (Tuesday)

Time:

3:30pm - 4:45pm

Venue:

Room 513, William M.W. Mong Engineering Building,
The Chinese University of Hong Kong

Abstract:

In automatic speech recognition systems, a crucial component is the language model (LM). In order to more robustly handle different styles or tasks, interpolated language models are often constructed by building multiple models on a collection of diverse text sources. These are to be combined using global interpolation weights. They may be tuned using either perplexity or discriminative approaches to adapt LMs to a particular task. However, the "usefulness" of sources can vary between different word contexts, and are determined by a number of factors, including modeling resolution, form of parameter estimation, topic and style.  Hence, the ability of using global interpolation to capture such local variability is poor. In this talk, a general form of context dependent model interpolation is introduced. Depending on the previous word contexts, a discrete history weighting function is used to dynamically adjust the contribution from each component model. Such a flexible form of model interpolation allows either a task  independent, or a target domain adapted language model to be discriminatively trained. An important part of this new approach is obtaining robust estimation. A number of schemes to address this issue are presented, MAP estimation based smoothing, using class based contexts and a simple back-off approach. Experimental evaluation was conducted on a state-of-the-art Mandarin broadcast transcription task. Consistent gains in perplexity using context dependent, rather than global, weights are observed as well as reductions in character error rate.

About the Speaker:

Dr. Andrew LIU is a Postdoctoral Research Associate in the Machine Intelligence Laboratory, Department of Engineering, Cambridge University. Dr. LIU graduated from Shanghai JiaoTong University in 2000 before pursuing an M.Phil. (2001) and PhD. (2006)  from Department of Engineering of Cambridge University. Dr. LIU completed his PhD on discriminative complexity control and linear projections in September 2005 under the supervision of Dr. Mark Gales. Dr. LIU's research interests include large-vocabulary continuous speech recognition, automatic model complexity control, discriminative training, language modeling and machine translation.

 

Light refreshment will be available after 3:15pm.

***** ALL ARE WELCOME *****

Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), +852-26098304

 

 


 

 Distinguished Lecture by Professor Victor ZUE 10/31/2008 12:00:00 AM 10/31/2008 12:00:00 AM

SHIAE & CUHK MoE-Microsoft Key Lab
 
Distinguished Lecture Series 2008
by
Professor Victor ZUE

We are delighted to have invited Professor Victor ZUE, Delta Electronics Professor of Electrical Engineering and Director of Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, USA, to visit The Chinese University of Hong Kong and deliver a distinguished lecture. Professor ZUE will share with us his personal opinion about the factors contributing to the innovation-rich research environment and his talk is entitled "Innovation: An MIT CSAIL Perspective". You are cordially invited to this lecture jointly organized by Shun Hing Institute of Advanced Engineering & CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies. The details are as follows:

DATE:              31 October, 2008 (Friday)
TIME:              4:45 p.m. - 6:15 p.m.
VENUE:           TY Wong Hall, 5/F, Ho Sin Hang Engineering Building, CUHK
SPEAKER:       Professor Victor ZUE, Massachusetts Institute of Technology, USA
TALK TITLE:  Innovation: An MIT CSAIL Perspective

 Abstracts
For more than four decades, the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and its predecessors ­ the Artificial Intelligence Laboratory and the Laboratory for Computer Science, have contributed many technical innovations, ranging from time-sharing and RSA public key encryption to robotics and human-like interfaces. Some of these innovations have spawned successful start-ups or have been acquired by multinationals. In this talk, I would like to offer my personal opinion about the factors contributing to this innovation-rich research environment. I will illustrate my points with examples drawn from past and current research.

Biography of Speaker
Victor ZUE received his ScD from MIT in 1976, and has been at MIT ever since. He is the Delta Electronics Professor of Electrical Engineering and the Director of the Computer Science and Artificial Intelligence Laboratory. In the early part of his career, Victor conducted research in acoustic-phonetic and phonological analyses of American English. Subsequently, his research interest shifted to the development of spoken language interfaces to make human-computer interactions easier and more natural. Between 1989 and 2001, he headed the Spoken Language Systems Group at the MIT Laboratory for Computer Science, which has pioneered the development of many systems that enable a user to interact with computers using spoken language.

Outside of MIT, Victor has served on many planning, advisory, and review committees for the U.S. federal government and for many multinational corporations. From 1996-1998, he chaired the Information Science and Technology Study Group for the Defense Advanced Research Projects Agency of the U.S. Department of Defense, helping the DoD formulate new directions for information technology research. In 1999, he received the DARPA Sustained Excellence Award. Victor is a Fellow of the Acoustical Society of America, a Fellow of the International Speech Communication Association, and a member of the U.S. National Academy of Engineering. He is also an Academician of Academia Sinica in Taiwan.

***** All Are Welcome  *****
* Light refreshment will be served at 4:30pm on 5/F *

Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), (852) 2609 8304

 

 


 

 Seminar on "When Multimedia Meets Internet: Challenges and Opportunities" 8/13/2008 12:00:00 AM 8/13/2008 12:00:00 AM

     **Seminar**

        CUHK MoE-Microsoft Key Laboratory of
        Human-Centric Computing and Interface Technologies
&
Department of Electronic Engineering

-------------------------------------------------------------------------------------------------------------

Title:                             When Multimedia Meets Internet: Challenges and Opportunities

Speaker: Dr. Shipeng Li
Principal Researcher and Research Manager of the Internet Media group
Microsoft Research Asia (MSRA)

Date: August 13, 2008 (Wednesday)

Time: 10:00a.m.

Venue:Room 222, Ho Sin-Hang Engineering Building,
The Chinese University of Hong Kong
Abstract: We’re in a fast paced internet era. More and more multimedia contents are put onto the web for people to share, distribute and search. The internet has aggregated an enormously large distributed multimedia database. How do we deal with such a Web-scale multimedia database to better serve people? What challenges and opportunities does it provide to us? What does it mean to multimedia research?
In this talk, I will use several projects that we are working on in the Internet Media group at Microsoft Research Asia as examples to share with you our vision and research philosophy on the internet-scale multimedia. Topics ranging from fundamentally more efficient media coding scheme, better web search for multimedia data, to web security will be addressed in my talk. Hopefully, this talk would inspire further discussion and collaborations in this area.

About the Speaker: Dr. Shipeng Li joined Microsoft Research Asia (MSRA) in May 1999. He is a principal researcher and research manager of the Internet Media group. He is now also acting managing the Visual Computing group. His research interests include Signal and Image Processing; Content-based Analysis; Image and Video Coding; HDTV Technology; Multimedia Streaming and Communications over Internet and Wireless Networks; Scalable Multimedia Representation; Application Level Multicast; Digital Right Management; Wireless Communication and Networking; P2P Networking; New Media Formats and Systems. From Oct. 1996 to May 1999, Dr. Li was with Multimedia Technology Laboratory at Sarnoff Corporation in Princeton, NJ (formerly David Sarnoff Research Center, and RCA Laboratories) as a member of technical staff. Dr. Li has been actively involved in research and development of digital television, MPEG, JPEG, image/video compression, next generation multimedia standards and applications, consumer electronics. He has made several major contributions adopted by MPEG-4, H.264, MPEG-4 SVC standards. He first proposed the 5/D Media 2.0 concepts that outlined the new features of next generation internet media (2006). He has authored and co-authored over 150 journal and conference papers, 7 books or book chapters, and holds 40 granted (and 40+ pending) US patents in image/video compression and communications, digital television, multimedia and wireless communication.
Dr. Li is the secretary of Visual Signal Processing and Communications technical committee of IEEE Circuits and Systems Society and a past member of Multimedia Signal Processing technical committee of IEEE Signal Processing Society. He serves as associate editors of IEEE Transactions on Circuits and Systems for Video Technology and Journal of Visual Communications and Image Representation. He has also served as chairs and/or members in over a dozen technical committees of various international conferences on multimedia, including ICME, VCIP, PV, PCM, & ISCAS.
Dr. Li is an adjunct professor of the Chinese University of Hong Kong. He is also PhD supervisors for University of Science and Technology of China and Shanghai Jiaotong University.
Dr. Li received his B.S. and M.S. in Electrical Engineering from the University of Science and Technology of China (USTC) in 1988 and 1991, respectively. He received his Ph.D. in Electrical Engineering from Lehigh University, Bethlehem, PA, in 1996.

***** ALL ARE WELCOME *****

Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), +852-26098304
              Senny Chan (pychan@ee.cuhk.edu.hk), +852-26098486

 


 

Distinguished Lecture by Professor Julia Hirschberg, Professor of Computer Science, Columbia University 3/11/2009 12:00:00 AM 3/11/2009 12:00:00 AM

CUHK MoE-Microsoft Key Lab
 
Distinguished Lecture Series 2009
by
Professor Julia Hirschberg

We are delighted to have invited Professor Julia Hirschberg, Professor of Computer Science, Columbia University, USA, to visit The Chinese University of Hong Kong and deliver a distinguished lecture. Professor Julia will discuss production and perception studies of deceptive speech with us and her talk is entitled "Acoustic, Prosodic and Lexical Cues to Deception in Speech". You are cordially invited to this lecture organized by CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies. The details are as follows:
 

 

DATE:             March 11, 2009 (Wednesday)
TIME: 4:30 p.m. - 5:30 p.m. 
VENUE: TY Wong Hall, 5/F, Ho Sin Hang Engineering Building, CUHK
SPEAKER: Professor Julia Hirschberg, Professor of Computer Science, Columbia University, USA
TALK TITLE: Acoustic, Prosodic and Lexical Cues to Deception in Speech

Abstracts
In this talk I will discuss production and perception studies of deceptive speech. We have collected a large corpus of deceptive and non-deceptive speech in order to identify acoustic/prosodic and lexical cues associated with deception. I will discuss machine learning experiments to predict deception which achieve performance which compares favorably with the performance of human judges in perception tests on the same data and task. We have also findings on the role of personality in the ability of human judges to detect deception.  

Biography of Speaker
Julia Hirschberg is Professor of Computer Science at Columbia University. From 1985-2003 she worked at Bell Labs and AT&T Labs, as member of Technical Staff working on intonation assignment in text-to-speech synthesis and then as Head of the Human Computer Interaction Research Department. Her research focuses on prosody in speech generation and understanding. She currently works on speech summarization, emotional speech, charismatic speech, deceptive speech, and dialogue prosody. Hirschberg was President of the International Speech Communication Association (ISCA) from 2005-2007 and co-editor-in-chief of Speech Communication from 2003-2006. She was editor-in-chief of Computational Linguistics and on the board of the Association for Computational Linguistics from 1993-2003.  She has been a fellow of the American Association for Artificial Intelligence since 1994 and an ISCA Fellow since 2008. 

 
***** All Are Welcome *****
 * Light refreshment will be served at 4:15pm *
 
Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), (852) 2609 8304

 

 


 

Seminar on New Results in Video Copy Detection Techniques by Professor C.C. Jay Kuo 3/20/2009 12:00:00 AM 3/20/2009 12:00:00 AM

**Seminar**
CUHK MoE-Microsoft Key Lab of Human-Centric Computing and Interface Technologies
------------------------------------------------------------------------------------------------

Title: New Results in Video Copy Detection Techniques
Speaker: Professor C.-C. Jay Kuo
University of Southern California
Date: March 20, 2009 (Friday)
Time: 2:15pm - 3:30pm
Venue: Room 513, William M.W. Mong Engineering Building, The Chinese University of Hong Kong

Abstract:
The rapid development of digital video processing technologies and the increasing bandwidth enable easy access, editing, and distribution of digital video contents. Copyright protection becomes a growing concern for content creators/owners nowadays. One of the key problems is duplicate video detection. For example, there are many illegal video copies upload to the YouTube website every day. It is important for YouTube to identify these illegal copies fast. In this work, we propose a novel video detection system that can identify duplicate video copies very efficiently. We propose a compact signature based on the underlying video structure, which is discriminative yet insensitive to various attacks. In addition, we propose to use an extremely efficient matching technique, which is originated from fast symbol string search. Unlike images and audio, the size of videos is usually very large, which makes it computationally expensive to match two very long video sequences. Our system can perform the matching in linear time while the computational cost usually grows at least quadratically as the length of the video for other existing solutions.

About the Speaker:
Dr. C.-C. Jay Kuo received the Ph.D. degrees from the Massachusetts Institute of Technology in 1987. He is now with the University of Southern California (USC) as Director of Signal and Image Processing Institute and Professor of EE, CS and Mathematics. His research interests are in the areas of digital media processing, multimedia compression, communication and networking technologies, and embedded multimedia system design. Dr. Kuo is a Fellow of IEEE and SPIE. Dr. Kuo has guided about 90 students to their Ph.D. degrees and supervised 20 postdoctoral research fellows. Currently, his research group at USC consists of around 35 Ph.D. students (see website http://viola.usc.edu), which is one of the largest academic research groups in multimedia technologies. He is a co-author of about 150 journal papers, 770 conference papers and 9 books. Dr. Kuo is Editor-in-Chief for the Journal of Visual Communication and Image Representation, and Editor for the Journal of Information Science and Engineering, LNCS Transactions on Data Hiding and Multimedia Security (a Springer journal), the Journal of Advances in Multimedia (a Hindawi journal) and the EURASIP Journal of Applied Signal Processing (a Hindawi journal).  He was on the Editorial Board of the IEEE Signal Processing Magazine in 2003-2004. He served as Associate Editor for IEEE Transactions on Image Processing in 1995-98, IEEE Transactions on Circuits and Systems for Video Technology in 1995-1997 and IEEE Transactions on Speech and Audio Processing in 2001-2003.
 

***** ALL ARE WELCOME *****
Light refreshment will be available after 2:00pm.

 


 

The Forth Beijing-Hong Kong International Doctoral Forum 8/6/2009 12:00:00 AM 8/10/2009 12:00:00 AM

The Forth Beijing-Hong Kong International Doctoral Forum 2009 on Network and Media Computing
Held at The Chinese University of Hong Kong, Hong Kong SAR, China
Please visit http://wiki.cse.cuhk.edu.hk/PHDForum/2009 for more information.

**Paper Submission Deadline extends to 30 June 2009**

*****************************************************************************************
                                                CALL FOR PAPERS

                    The Fourth Beijing-Hong Kong International Doctoral Forum 2009

Hong Kong, China
6-10 August 2009
http://wiki.cse.cuhk.edu.hk/PHDForum/2009

Beijing-Hong Kong Doctoral Forum 2009 invites you to experience both the vibrant and cosmopolitan atmosphere of Hong Kong while enjoying the high quality forum programs. The objective of the forum is to bring together Ph.D. students from all over the world, contributing to all facets of networking and multimedia computing.  All expenses within the forum of the attendees will be covered by CUHK, including registration, meals, accommodation, and local transportation fees, etc. 

****************************************
IMPORTANT DATES:

  *  Paper submission: June 15, 2009 (4 pages IEEE format)
  *  Notification of acceptance: June 30, 2009
  *  Registration: July 6, 2009
  *  Camera Ready Due: July 6, 2009
  *  Forum: August 6-10, 2009

****************************************
AWARDS:

Best paper, Best Poster, Best Presentation and Best Demonstration awards will be selected and announced during the forum.

****************************************
 
EVENT HIGHLIGHTS:
  *  Invited talks
  *  Oral presentations and poster sessions
  *  Research demonstrations
  *  Social activities
  *  Local tours

****************************************
PAPER SUBMISSION:

Student authors are invited to submit research papers on original work (including recently published work) describing current research and novel ideas in Network Modeling and Design, Media Computing, Systems and Applications, Web Intelligence and Social Computing. Papers whose contributions are supported by experimental evaluations are strongly encouraged.

Copyright of the submitted work is held by the author/owner.

Authors should submit a four-page manuscript in double-column format including authors' names, affiliations, and a short abstract. Recently published conference papers are also welcome. Only electronic submission written in English will be accepted. Please download required paper format and sample files at the forum website:
  http://wiki.cse.cuhk.edu.hk/PHDForum/2009/author_s_kit

Submissions can be made in the following link (Please create an EASYCHAIR account to submit):
  http://www.easychair.org/conferences/?conf=idf2009

Papers are solicited in all areas of network and media computing, including, but not limited to:

Network Modeling and Design:
 - Network architectures and algorithms
 - Network security
 - Network quality, reliability and fault-tolerance
 - Home, mobile, wireless, ad-hoc and broadband networks
 - Next generation networks

Media Computing:
 - Multimedia coding and trans-coding
 - Audio/video streaming
 - Multimedia content analysis and indexing
 - Multimedia authoring and retrieval
 - Multi-modal human-computer interaction and integration
 - Multi-modal biometric authorization
 - Multimedia security and rights management
 - Signal Processing and machine learning

Systems and Applications:
 - System performance measurement and benchmark
 - Multimedia search engines
 - Multimedia in e-learning, entertainment and virtual reality
 - Telecommunication

Web Intelligence and Social Computing:
 - Web structural, content, and usage mining
 - Web information clustering and classification
 - Web page search, crawling and ranking techniques
 - Web metrics and statistics
 - Social network modeling and analysis
 - Collaborative filtering and recommender systems
 - Social media systems like wikis, blogs, micro-blogs, mashup, etc.
 - Social ranking, searching, and large-graph algorithms

****************************************
ORGANIZERS:
  *  The Chinese University of Hong Kong
  *  Tsinghua University

ORGANIZATION:

-Professors-
Conference General Co-chairs
  *  Helen Meng, The Chinese University of Hong Kong
  *  Irwin King, The Chinese University of Hong Kong

Organizing Committee
  *  Jeffrey Yu, The Chinese University of Hong Kong
  *  Hong Cheng, The Chinese University of Hong Kong
  *  Jiaya Jia, The Chinese University of Hong Kong
  *  Yongjin Liu, Tsinghua University
  *  Guoliang Li, Tsinghua University

Advisory Committee
  *  Maosong Sun, Tsinghua University
  *  Juanzi Li, Tsinghua University
  *  Peter Yum, The Chinese University of Hong Kong

-Students-
Organizing Chair and Co-chair
  *  Josh Lam, The Chinese University of Hong Kong
  *  Moses Lee, The Chinese University of Hong Kong

Programming Chair and Co-chair
  *  Zhang Wei, The Chinese University of Hong Kong
  *  Yang Xuan, The Chinese University of Hong Kong

****************************************
SPONSORS:
  *  CUHK MoE-Microsoft Key Laboratory for Human-Centric Computing and Interface Technologies?c   
  *  Microsoft Research Asia

****************************************
 

 


 

Seminar on Two and Three-Dimensional Visual Articulatory Models for Pronunciation Training and for Treatment of Speech Disorders 6/2/2009 12:00:00 AM 6/2/2009 12:00:00 AM

**Seminar**
CUHK MoE-Microsoft Key Laboratory of

        Human-Centric Computing and Interface Technologies

------------------------------------------------------------------------------------------------

Title:                            

Two and Three-Dimensional Visual Articulatory Models for Pronunciation Training and for Treatment of Speech Disorders

Speaker:

Professor Bernd J. Kröger
Aachen Technical University (RWTH)
Germany

Date:

June 2, 2009 (Tuesday)

Time:

11:00am - 12:15pm

Venue:

Room 513, William M.W. Mong Engineering Building,
The Chinese University of Hong Kong

Abstract:

Visual articulatory models can be used for visualizing vocal tract articulatory speech movements. This information may be helpful in pronunciation training or in therapy of speech dis­orders. For testing this hypothesis, speech recog­nition rates were quantified for mute animations of vocalic and conso­nantal speech movements generated by a 2D and a 3D visual articulatory model. The visually based speech sound recog­nition test (mimicry test) was performed by two groups of eight children (five to eight years old) matched in age and sex. The children were asked to mimic the visually produced mute speech movement anima­tions for different speech sounds. Recog­nition rates stay signifi­cantly above chance but indicate no significant difference for each of both models. Children older than 5 years are capable of interpreting vocal tract articulatory speech sound move­ments without any preparatory training in a speech adequate way. The complex 3D-display of vocal tract articulatory mo­ve­­ments provides no significant advan­tage in comparison to the visually simpler 2D-midsagittal displays of vocal tract arti­cu­­la­tory movements.
 

About the Speaker:

Bernd Kröger studied physics at University of Münster (Germany) and got his masters degree in physics in 1985. He studied Phonetics and Music Science at University of Cologne (Germany) and got his PhD degree in 1989 in phonetics. He got his Postdoctoral Lecture Qualification ("Habilitation") in Phonetics, At University of Cologne in 1998. Bernd Kröger was research associate at the Department of Phonetics (Head: Prof. Dr. G. Heike), University of Cologne from 1987-1990 and Assistant Professor at the same Department from 1990 to 1998. Bernd Kröger was Guest Professor at the Centre for General Linguistics, Typology and Universals Research (ZAS) Berlin and at the Department for German Language and Linguistics, Humboldt University of Berlin, Germany from 1999-2000. From 2001 til now Bernd Kröger is senior scientific staff memberand professor at the Department of Phoniatrics, Pedaudiology, and Communication Disorders, UKA and RWTH Aachen, Germany. Bernd Kröger was leader of over 10 research projects funded by the German Research Council of by the European Government. His main research subject is speech production, speech acoustics and speech synthesis.

Light refreshments will be available after 10:45am.

 ***** ALL ARE WELCOME *****
Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), +852-26098304

 


 

The Plaque Unveiling Ceremony of CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies 8/7/2009 12:00:00 AM 8/7/2009 12:00:00 AM

The Plaque Unveiling Ceremony of
CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and
Interface Technologies

Date: August 7, 2009 (Friday)
Time: 11am
Venue: TY Wong Hall, 5/F, Ho Sin Hang Engineering Building, CUHK

Inquiries: Miss Linda Ma (tsma@se.cuhk.edu.hk) or (852) 26098304

 


 

View-Upload Decoupling: A Redesign of Multi-Channel P2P Video Systems 7/21/2009 12:00:00 AM 7/21/2009 12:00:00 AM

CUHK MoE-Microsoft Key Lab of
Human-Centric Computing and Interface Technologies 
 
Distinguished Lecture Series 2009

by
Professor Keith W. Ross

 

We are delighted to have invited Professor Keith W. Ross from Polytechnic Institute of New York University of USA to visit The Chinese University of Hong Kong and deliver a distinguished lecture. Professor Ross will present infinite-server queuing network models to analytically study the performance of multi-channel P2P streaming systems and his talk, "View-Upload Decoupling: A Redesign of Multi-Channel P2P Video Systems", is from the paper that won the Best Paper Award in Infocom 2009. You are cordially invited to this lecture jointly organized by Department of Information Engineering & CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies. The details are as follows:
DATE:             21 July, 2009 (Tuesday)
TIME: 11:00 a.m. - 12:00 p.m. 
VENUE: Room 1009, Engineering Building II, CUHK
SPEAKER: Professor Keith W. Ross, Polytechnic Institute of New York University, USA
TALK TITLE: View-Upload Decoupling: A Redesign of Multi-Channel P2P Video Systems

          

 

                 
Abstracts
In recent years there have been several large-scale deployments of P2P live video systems. Existing and future P2P live video systems will offer a large number of channels, with users switching frequently among the channels. In this talk, we present infinite-server queuing network models to analytically study the performance of multi-channel P2P streaming systems. Our models capture the essential aspects of multi-channel video systems, including peer channel switching, peer churn, peer bandwidth heterogeneity, and Zipf-like channel popularity. We apply the queuing network models to two P2P streaming designs: isolated channel design (ISO) and View-Upload Decoupling (VUD) design. For both of these designs, we develop efficient algorithms to calculate critical performance measures, develop an asymptotic theory to provide closed-form results when the number of peers approaches infinity, and derive near-optimal provisioning rules for assigning peers to groups in VUD. 
 
Biography of Speaker
Professor Ross joined Polytechnic University as the Leonard J. Shustek Chair Professor in Computer Science in January 2003. He has been Department Head since September 2008. Before joining Polytechnic, he was a professor for five years in the Multimedia Communications Department at Eurecom Institute in Sophia Antipolis, France. From 1985 through 1997, he was a professor in the Department of Systems Engineering at the University of Pennsylvania. He received a B.S.E.E from Tufts University, a M.S.E.E. from Columbia University, and a Ph.D. in Computer and Control Engineering from The University of Michigan.

Professor Ross has worked in peer-to-peer networking, Internet measurement, video streaming, Web caching, multi-service loss networks, content distribution networks, network security, voice over IP, optimization, queuing theory, and Markov decision processes. He is an IEEE Fellow, and is currently associate editor for IEEE/ACM Transactions on Networking. He has served as an advisor to the Federal Trade Commission on P2P file sharing.

Professor Ross is co-author (with James F. Kurose) of the popular textbook, Computer Networking: A Top-Down Approach Featuring the Internet, published by Addison-Wesley (first edition in 2000, fifth edition 2009). It is the most popular textbook on computer networks in CS departments, both nationally and internationally; it has been translated into twelve languages. Professor Ross is also the author of the research monograph, Multiservice Loss Models for Broadband Communication Networks, published by Springer in 1995.

>From July 1999 to July 2001, Professor Ross took a leave of absence to found and lead Wimba, an Internet technology start-up. Wimba develops and markets Java-based asynchronous and synchronous voice-over-IP technologies, primarily for the on-line education and language learning markets. Wimba is now headquartered in NYC and has more than 80 employees worldwide.
 

 

==============================
Department of Information Engineering
CUHK MoE-Microsoft Key Laboratory
TEL:(852)2609 8385 / (852)2609 8304
Website: http://www.ie.cuhk.edu.hk and www.cuhk.edu.hk/ms-cu-jl
Email: dept@ie.cuhk.edu.hk and ms-cu-jl@se.cuhk.edu.hk
==============================

**** All Are Welcome  *****

For more details, please visit http://www.cuhk.edu.hk/ms-cu-jl/
Thank you.

 


 

Seminar by Professor James A. Landay, University of Washington 10/2/2009 12:00:00 AM 10/2/2009 12:00:00 AM

 

**Seminar**
CUHK MoE-Microsoft Key Lab of

        Human-Centric Computing and Interface Technologies
------------------------------------------------------------------------------------------------

Title:                            
Activity-based UbiComp: A New Research Basis for the Future of Human-Computer Interaction
 
Speaker:
Professor James A. Landay
Computer Science & Engineering
University of Washington
 
Date:
October 2, 2009 (Friday)
Time:
4:00pm - 5:30pm
Venue:
Room 703, William M.W. Mong Engineering Building,
The Chinese University of Hong Kong
Abstract:
There are many indicators that people feel technology is speeding up and complicating their lives. The Digital Simplicity project instead offers a different value proposition: take this simple technology into your life and we will make the high-level, long-lived activities you are currently doing easier. We apply novel technology in physical activity inference and mobile device UIs to support people’s high-level, long-lived activities. Target activities include supporting elders in staying healthy and independent, motivating individuals to get fit, and reducing a family’s environmental footprint. These types of applications require a new design methodology as well as a new set of tools to make it easier to collect design information, prototype applications, and evaluate design ideas in situ. We will demonstrate activity-based applications we have designed as well as novel tools for prototyping mobile, activity-based applications and collecting in-situ data on mobile phones.
About the Speaker:
James Landay is an Associate Professor in Computer Science & Engineering at the University of Washington, specializing in human-computer interaction. From 2003 through 2006 he was also the Laboratory Director of Intel Research Seattle, a university affiliated research lab exploring ubiquitous computing. His current research interests include Automated Usability Evaluation, Demonstrational Interfaces, Mobile & Ubiquitous Computing, User Interface Design Tools, and Web Design.  He is spending his 2009-2010 sabbatical at Microsoft Research Asia in Beijing. Landay received his B.S. in EECS from UC Berkeley in 1990 and M.S. and Ph.D. in CS from Carnegie Mellon University in 1993 and 1996, respectively. His Ph.D. dissertation was the first to demonstrate the use of sketching in user interface design tools. He was also the chief scientist and co-founder of NetRaker. In 1997 he joined the faculty in EECS at UC Berkeley, leaving as an Associate Professor in 2003.
 

***** ALL ARE WELCOME *****
 
Light refreshment will be available after 3:45pm.
 
Inquiries: Linda Ma (tsma@se.cuhk.edu.hk), +852-26098304

 

=============================

Attendance taken:

  • Wang Wei (CSE)
  • Liu Xueting (CSE)
  • Dai Zhenlong (CSE)
  • Jiang Lei (CSE)

 


 

Public Lecture on Rethinking Computing by Mr. Craig Mundie jointly organized by Microsoft and CUHK 11/16/2009 12:00:00 AM 11/16/2009 12:00:00 AM

register

 

Speech Title:
    Rethinking Computing
Speaker: 
    Craig Mundie
    Chief Research and Strategy Officer
    Microsoft Corporation
Speech abstract:
    Please join us for a special presentation by Craig Mundie, chief research and strategy
    officer of Microsoft. Craig is one of two senior executives who took over from Bill Gates.
    Craig will talk about the technologies that will change how people interact with the world,
    and the opportunities for every academic discipline to harness the power of computing.
    Craig will showcase a number of advanced technologies, and answer questions from the
    audience.
Date:
    November 16, 2009, Monday
Time:
    4:00 – 5:30 pm
    (registration starts at 3:35 pm)
Venue:
    Lecture Theatre 6 (LT-6)
    Teaching Complex at Western Campus
    The Chinese University of Hong Kong
    Shatin, N.T.

Craig’s bio and picture can be found at: http://www.microsoft.com/presspass/exec/craig/

 


 

Seminar on "Knowledge is Power in Data-Driven NLP" by Dr. Ye-Yi Wang 11/9/2009 12:00:00 AM 11/9/2009 12:00:00 AM

**Seminar**
CUHK MoE-Microsoft Key Lab of
Human-Centric Computing and Interface Technologies

Title:                             Knowledge is Power in Data-Driven NLP

Speaker: Dr. Ye-Yi Wang
Researcher
Microsoft Redmond

Date: November 9, 2009 (Monday)
Time: 3:00pm - 4:15pm
Venue:Room 513, William M.W. Mong Engineering Building,
          The Chinese University of Hong Kong
 

Abstract: The theories of language acquisition are often categorized into two camps – the rational approach states that most of the human language faculty is innate, while the empirical approach stresses that the languages are mostly acquired through external experience. Reflected in the field of NLP, we have seen the knowledge-based approach and the data-driven approach. This simple categorization ignores the facts that the rationalists never deny the effect of external experience and the empiricists also admit some basic hardwired rules in human brain. In case of NLP, it also makes sense to integrate the knowledge-based and the data-driven approaches. For over a decade, we have been working on different aspects of data-driven NLP, including statistical MT, language modeling, spoken language understanding, and information extraction. A common lesson learned from all the project is the power of knowledge – it helped either improve the performance or reduce the requirement of training data significantly. In this talk I will review some of the projects and show in details how knowledge can be learned and integrated in data-driven NLP systems.

About the Speaker: Ye-Yi Wang received a B.S. and a M.S. degree in Computer Science from Shanghai Jiao Tong University, a Master’s degree in Computational Linguistics from Carnegie Mellon University and a Ph.D. degree in Human Language Technology from Carnegie Mellon University. Dr. Wang joined Microsoft Research (Redmond, WA) in 1998. His research interests include user intent understanding, spoken dialog systems, natural language processing, information retrieval, language modeling, statistical machine translation and machine learning. He served in the editorial board of the Chinese Contemporary Linguistic Theory Series. He is a co-author of Introduction to Computational Linguistics (China Social Sciences Publishing House, 1997), and he has published over 50 journal and conference papers. Dr. Wang is a member of ACL and ACM, and a senior member of IEEE.

***** ALL ARE WELCOME *****

Light refreshment will be available after 3:45pm.

 


 

Seminar on "Talking "Interaction" Seriously:How to Engage Japanese Learners of English in Oral and/or Written Communication in College Language Classes 3/23/2010 12:00:00 AM 3/23/2010 12:00:00 AM

**Seminar**

CUHK MoE-Microsoft Key Lab & Childhood Bilingualism Research Centre

------------------------------------------------------------------------------------------------

Title: Taking "Interaction" Seriously: how to engage Japanese learners of English
         in oral and/or written communication in college language classes

Speaker: Professor Yasunari HARADA
                 Professor of Faculty of Law 
                 Director of Institute for Digital Enhancement of Cognitive Development 
                 Waseda University

Date: March 23, 2010 (Tuesday)

Time: 4:30pm - 5:30pm

Venue:Swire Hall 2, Fung King Hey Building

Abstract:
College English teachers in Japan face several obstacles when trying to get their students engage in communication in their language classes. For most high school students in Japan, their central objective in learning English is to get better scores in college entrance examinations. Upon entrance, they have lost their greatest motivation for continuing to learn English seriously. Throughout their school life from kindergarten to preparatory (cram) schools, they are taught to sit silent and obedient in class.  Anyone who asks questions or express their opinions, especially those that are not expected by the teachers, are severely penalized sooner or later. In their Japanese classes, they have been taught to repeat the main-stream ideas and not to be original and creative, or logical and argumentative, or imaginative and persuasive. College English teachers have to organize their classes in such a way that those students unlearn most important things they have acquired during their 12 or more years of their school experiences, most importantly, in the way they deal with the English language, the way they interact with other students, and most importantly, the way they view what learning English is all about. In this talk, I will first present those problems and give you an overview of some of the activities in my English classes and go on to the data collection efforts that is going on for four years now and will continue for another four-year period, in which we gather and store electronic documents that those students create and revise during class and as homework plus audio and video recordings of those activities. On the one hand, those collections of data would form a basis for written and spoken learner corpus, from which we can extract linguistic characteristics of those learners' production in English. However, we will also be able to extract important insight into the way students motivate and help each other by engaging in a real-life communication, in English and Japanese. For instance, in most language, liberal arts and law courses offered in the undergraduate School of Law, where I teach, an average student would attend the classes regularly without making even a single new friend, which never happens in my classes. By engaging in various activities, in constantly changing groupings, students naturally get to know each other well, and by the end of the term, they will have some good friends and many new acquaintances. They also learn that speaking and writing in English with their peer classmate in mind, is fun rather than drudgery. Expressing their ideas and opinions, with concrete examples and/or reasons, become their second nature. And most if not all of those students learn that getting a better score is not all there is to learning English.

About the Speaker:
Yasunari HARADA is a Professor at Faculty of Law and Director at the Institute for Digital Enhancement of Cognitive Development at the Waseda University. Yasunari is also a President of Logico-Linguistics Society of Japan, as well as a PACLIC Steering Committee member. You may know more  details about Yasunari at
 http://www.f.waseda.jp/harada/profile09-e.html

***** ALL ARE WELCOME *****

Light refreshment will be available after 4:15pm.

POSTER

 


 

Temporal Data Clustering Different Representations by Dr. Ke Chen 4/16/2010 12:00:00 AM 4/16/2010 12:00:00 AM

Title:
Temporal Data Clustering Different Representations

Speaker:
Dr. Ke Chen
Senior Lecturer
Department of the Computer Science
The University of Manchester

Date:  
April 16, 2010 (Friday)

Time: 
4:30 – 5:30 p.m.

Venue:  
Room 906, William M.W. Mong Engineering Building,
The Chinese University of Hong Kong

Abstract:  
Temporal data clustering algorithms provides underpinning techniques for discovering the intrinsic structure and condensing/summarizing information conveyed in temporal data. Such techniques are demanded in various fields ranging from time series analysis to sequential data understanding. In this talk, I shall present a novel temporal data clustering framework via weighted clustering ensemble on different temporal data representations. I will begin with introducing the background and our motivation, which leads to a novel approach to temporal data clustering. I shall describe our temporal data clustering framework along with the critical supporting technique -- our weighted clustering ensemble algorithm in detail. Then I report experimental results in a variety of temporal data clustering tasks, including time series benchmark data mining, motion trajectory clustering analysis and temporal data stream clustering, to demonstrate the effectiveness and flexibility of our approach in temporal data clustering. Finally I shall discuss some related issues and present conclusion remarks.

About the Speaker:  
Ke Chen received the BSc, MSc and PhD degrees in 1984, 1987 and 1990, respectively, all in computer science. He has been with The University of Manchester since 2003. He was with The University of Birmingham, Peking University, The Ohio State University, Kyushu Institute of Technology, and Tsinghua University. He was a visiting professor at Microsoft Research Asia in 2000 and Hong Kong Polytechnic University in 2001. He has been on the editorial board of several academic journals including IEEE Transactions on Neural Networks and serves as the category editor of Machine Learning and Pattern Recognition in Scholarpedia. He was the program chair of the first International Conference on Natural Computation and has been a member of the technical program committee of numerous international conferences including CogSci and IJCNN. He chairs Intelligent Systems Applications Technical Committee (ISATC) and University Curricula Subcommittee, IEEE Computational Intelligence Society. He also served as task force chairs and a member of NNTC, ETTC and DMTC in IEEE CIS. He is a recipient of several academic awards including the NSFC Distinguished Principal Young Investigator Award and JSPS Research Award. He has published over 100 academic papers in refereed journals and conferences. His current research interests include pattern recognition, machine learning, machine perception and computational cognitive systems.

****** ALL ARE WELCOME ***** *
Light refreshment will be available after 4:15pm.
Tel: 2609-8304 for enquiries

 


 

The Fifth Beijing-Hong Kong International Doctoral Forum 8/23/2010 12:00:00 AM 8/25/2010 12:00:00 AM

The 5th BJ-HK International Doctoral Forum 2010

 

Cloud computing, Social Networks, Computer Networks, Data Management and Rich Media Computing

August 23-25, 2010
Beijing, China

 

The 5th Beijing-Hong Kong International Doctoral Forum-2010 invites you to experience both the vibrant and cosmopolitan atmosphere of Beijing while enjoying the high quality forum programme. The objective is to bring together Ph.D. students from all over the world, contributing to all facets of the web, network and rich media computing etc. Expenses within the forum for the attendees will be covered by Tsinghua University. This includes registration, meals, accommodation, local transport, etc.
The Forum-2010 will be hosted by Tsinghua University, Beijing.
 

 

IMPORTANT DATES:
- Paper submission deadline July 15, 2010 (4 pages IEEE format)
- Notification of acceptance July 30, 2010
- Registration August 10, 2010
- Camera-ready paper due date August. 15, 2010
- Forum August 23-25, 2010
 

Please visit http://wiki.cse.cuhk.edu.hk/PHDForum/2010/start for more information

 


 

Seminar on Dolby's research with a daily impact by Dr. Claus Bauer 7/6/2010 12:00:00 AM 7/6/2010 12:00:00 AM

Dolby's research with a daily impact
by
Claus Bauer, Director Research, Dolby Laboratories
3:00 - 4:00 pm, July 6, 2010 (Tuesday)
SHIAE conference room

Since its early days in the 1960's, Dolby's invention have continuously changed and enhanced the way people all over the world enjoy music. Initially led by Dolby's founder Ray Dolby, Dolby?s Research department has always been and continues to be instrumental in inventing new audio technologies. This talk describes through examples how Dolby Research has invented technologies that were integrated in billions of consumer electronics devices, personal computers, mobile phones, and portable music players. It explains how algorithm design, statistics, and digital signal processing are used to define the ways we enjoy music.

Biography of Claus Bauer
In 1996, Claus Bauer received his Ph.D. degree in mathematics from the University of Freiburg, Germany. For his dissertation in number theory, he spent three years at universities in Beijing and Jinan, China. From 1997 -2000, he worked at Siemens Mobile, Munich on network design algorithm for GSM networks and system architecture for 3/4G cellular systems. From 2000- 2002, he was the lead researcher for IP data networking research at the Tellabs Research Center, Cambridge, Massachusetts. His research covered optical switching, media gateways, and quality of service technologies. Claus has been with Dolby, San Francisco since Sep. 2002. His research focuses on home networking technologies, mathematical modeling and optimization, theoretical analysis of coding techniques, error coding, and multimedia fingerprinting. As Director of Research, he is currently responsible for some of Dolby's global research initiatives. He has published over 50 papers in mathematics, networking, and multimedia processing, and speaks seven languages.

NULL 0 1 103 1 SEMINAR 10/5/2010 12:38:51 PM 10/5/2010 12:38:51 PM Seminar on Speech Prosody �V the Missing Link between Spoken and Written Language 10/12/2010 12:00:00 AM 10/12/2010 12:00:00 AM

CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies
&
Department of Linguistics and Modern Languages

Seminar

Date: October 12, 2010 (Tuesday)
Time: 4:30 p.m. – 6:15 p.m.
Venue: LT4, Teaching Complex at Western Campus (TCW)

Title: Speech Prosody – the Missing Link between Spoken and Written Language
Speaker: Professor Daniel Hirst
         Director de Recherches at the CNRS laboratory Parole et Langage
         University of Provence

Abstract:
In this presentation I argue that speech prosody is the crucial link between spoken and written language. When we speak, HOW we say something is even more important than WHAT we say. I will present recent results from recent work on modelling speech rhythm and speech melody using a large database of English.

Biography:
Daniel Hirst is a linguist and phonetician, who lives and works in the South of France. He has been working in the field of speech prosody and phonology for nearly forty years. He completed a doctoral thesis in 1974 on the role of intonation in the disambiguation of syntax. This was published by Mouton in the Janua Linguarum series (Hirst 1977). He completed a second thesis (Doctorat d'Etat - Habilitation) in 1987 on the linguistic description of prosodic systems. He is at present Directeur de Recherches at the CNRS laboratory Parole et Langage in the University of Provence, Aixen-Provence, where he co-directs a research team entitled Linguistic models, annotation and interfaces . Daniel Hirst has published numerous articles in several major journals (including Linguistics, Phonetica, Journal of Semantics, Phonology, Mind and Language, Linguistic Inquiry, Speech Communication ) and has contributed hapters to several international collaborative volumes. He was responsible for editing the first book on non-linear phonology to be published in France (Dell, Hirst & Vergnaud 1984).and for the edition of a major study of the intonation of languages of the world (Hirst & Di Cristo (eds)1998) to which he contributed the chapter on British English (Hirst 1998) and an 80 page introduction (Hirst & DiCristo 1998) in which he proposed a new international transcription system for intonation (INTSINT).

***** ALL ARE WELCOME *****

Enquiry: Miss Alice Lau (26098304, kylau@se.cuhk.edu.hk)
 

 


 

Tutorial on Analysing Speech Prosody with Momel and Intsint 10/15/2010 12:00:00 AM 10/15/2010 12:00:00 AM

CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies
&
Department of Linguistics and Modern Languages

Tutorial

Date: October 15, 2010 (Thursday)
Time: 3:30 p.m. – 5:15 p.m.
Venue: Rm 410L, Teaching Complex at Western Campus (TCW)

Title: Analysing Speech Prosody with Momel and Intsint
Speaker: Professor Daniel Hirst
         Director de Recherches at the CNRS laboratory Parole et Langage
         University of Provence

Abstract:
The tutorial will demonstrate the use of Praat, a free open source speech editor, for the analysis of speech prosody. he Momel and Intsint algorithms which I will describe have been successfully applied to a number of languages, including recently to spontaneous speech in standard Chinese (Beijing). A plugin application for the Momel and Intsint algorithms will be demonstrated together with the possibility of using Praat for the re-synthesis of speech from an abstract representation in order to test different phonological models.

Biography:
Daniel Hirst is a linguist and phonetician, who lives and works in the South of France. He has been working in the field of speech prosody and phonology for nearly forty years. He completed a doctoral thesis in 1974 on the role of intonation in the disambiguation of syntax. This was published by Mouton in the Janua Linguarum series (Hirst 1977). He completed a second thesis (Doctorat d'Etat - Habilitation) in 1987 on the linguistic description of prosodic systems. He is at present Directeur de Recherches at the CNRS laboratory Parole et Langage in the University of Provence, Aixen-Provence, where he co-directs a research team entitled Linguistic models, annotation and interfaces . Daniel Hirst has published numerous articles in several major journals (including Linguistics, Phonetica, Journal of Semantics, Phonology, Mind and Language, Linguistic Inquiry, Speech Communication ) and has contributed hapters to several international collaborative volumes. He was responsible for editing the first book on non-linear phonology to be published in France (Dell, Hirst & Vergnaud 1984).and for the edition of a major study of the intonation of languages of the world (Hirst & Di Cristo (eds)1998) to which he contributed the chapter on British English (Hirst 1998) and an 80 page introduction (Hirst & DiCristo 1998) in which he proposed a new international transcription system for intonation (INTSINT).

***** ALL ARE WELCOME *****

Enquiry: Miss Alice Lau (26098304, kylau@se.cuhk.edu.hk)
 

 


 

Seminar by Professor Shrijanth Narayanan on Enriched Spoken Language Processing and Applications 10/29/2010 12:00:00 AM 10/29/2010 12:00:00 AM

Seminar by Professor Shrijanth Narayanan on Enriched Spoken Language Processing and Applications

 

Title: Enriched Spoken Language Processing and Applications

Speaker: Professor Shrikanth (Shri) Narayanan

    University of Southern California

    University of Southern California Viterbi School of Engineering, Los Angeles, CA

 

Date: 29 October 2010 (Friday)

Time: 10:00 am -11:00 am

Venue: 513 William M. W. Mong Engineering Building (ERB513)

 

 Abstract:

The human speech signal is unique in the sense that it carries crucial information about not only communication intent and speaker identity but also underlying expressions and emotions. Automatically processing and decoding spoken language hence is a vastly challenging, and an inherently interdisciplinary, multifaceted endeavor. Recent technological approaches that have leveraged judicious use of both data and knowledge have yielded significant advances in this regards--beyond merely extracting underlying lexical information using automatic speech to text transcription--especially in terms of deriving rich information about prosody, discourse, and affect.   This talk will focus on some of the advances and open challenges in creating algorithms for machine processing of spoken language including their applications in areas such as enriched speech translation and behavioral informatics.  

 

 

 

Biography of the Speaker:

 

Shrikanth (Shri) Narayanan is the Andrew J. Viterbi Professor of Engineering at the University of Southern California (USC), where he holds appointments as Professor of Electrical Engineering, Computer Science, Linguistics and Psychology, and as Director of the USC Ming Hsieh Institute. Prior to USC he was with AT&T Bell Labs and AT&T Research. His research focuses on human-centered information processing and communication technologies.  Shri Narayanan is a Fellow of the Acoustical Society of America, IEEE, and the American Association for the Advancement of Science (AAAS).  He is an Editor for the Computer Speech and Language Journal and an Associate Editor for the IEEE Transactions on Multimedia, IEEE Transactions on Affective Computing and the Journal of the Acoustical Society of America. He He is a recipient of several awards including Best Paper awards from the IEEE Signal Processing society in 2005 (with Alex Potamianos) and in 2009 (with Chul Min Lee) and selection as a Distinguished Lecturer for the IEEE Signal Processing society for 2010-11.

Enquiry: Miss Alice Lau (kylau@se.cuhk.edu.hk)

 


 

Distinguished Lecture on The Invisible Computing Project by Professor Raj Reddy 11/8/2010 12:00:00 AM 11/8/2010 12:00:00 AM

Distinguished Lecture Jointly Organized by
Shun Hing Institute of Advanced Engineering and
CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies

Title: The Invisible Computing Project
Speaker: Professor Raj Reddy
Mozah Bint Nasser University Professor of Computer Science and Robotics,
School of Computer Science, Carnegie Mellon University, USA

Date: 8 November 2010 (Monday)
Time: 11:15 - 12:45
Venue: TY Wong Hall, 5/F Ho Sin Hang Engineering Building

Abstracts:
During the late 90s an ambitious project called “Invisible Computing” was launched at CMU to transform the way we use information technology under support from DARPA. It was discontinued for non-technical reasons.  It was intended to be a natural successor to Ubiquitous Computing and Pervasive Computing Concepts.  It had the characteristics of involving a department wide participation.  In this talk I will describe the Vision and the Plan for a Prototype Demonstration of the project.

Biography of the Speaker:
Dr. Raj Reddy is the Mozah Bint Nasser University Professor of Computer Science and Robotics in the School of Computer Science at Carnegie Mellon University.  From 1960-63, Dr. Reddy worked as an Applied Science Representative for IBM Corp., in Australia. He was an Assistant Professor of Computer Science at Stanford from 1966-69. He joined the Carnegie Mellon faculty as an Associate Professor of Computer Science in 1969. He became a Full Professor in 1973, and a University Professor in 1984. He served as the founding Director of the Robotics Institute from 1979 to 1991 and as the Dean of School of Computer Science from 1991 to 1999.
Dr. Reddy received a BE degree from the Guindy Engineering College of the University of Madras, India in 1958 and a MTech degree from the University of New South Wales, Australia, in 1960. He received a Ph.D. degree in Computer Science from Stanford University in 1966. Dr. Reddy's research interests include the study of human-computer interaction and artificial intelligence. His current research interests include Million Book Digital Library Project; Fiber To The Village Project; and Learning by Doing.
His professional honors include: Fellow of the Institute of Electrical and Electronics Engineers, Fellow of the Acoustical Society of America, Fellow of the American Association for Artificial Intelligence, Member of the National Academy of Engineering and Member of the American Academy of Arts and Sciences. He was president of the American Association for Artificial Intelligence from 1987 to 89. He is a recipient of the IBM Research Ralph Gomory Fellow Award in 1991.
Dr. Reddy was awarded the Legion of Honor by President Mitterand of France in 1984 and Padma Bhushan by President of India in 2001. He was awarded the ACM Turing Award in 1994, the Okawa Prize in 2004, the Honda Prize in 2005, and the Vannevar Bush Award in 2006. He served as co-chair of the President's Information Technology Advisory Committee (PITAC) from 1999 to 2001. He has been awarded honorary doctorates (Doctor Honoris Causa) from SV University, Universite Henri-Poincare, University of New South Wales, Jawaharlal Nehru Technological University, University of Massachusetts, University of Warwick, Anna University and the Indian Institute for Information Technology (Allahabad), Andhra University and IIT Kharagpur.  He serves on the Board of Governors of Peres Institute for Peace in Israel.

~~~ ALL ARE WELCOME ~~~

For enquiry: +852-31634351, info@shiae.cuhk.edu.hk

 


 

Distinguished Lecture on Crowds and Clouds for Spoken Language Processing by Professor Jim Glass 11/22/2010 12:00:00 AM 11/22/2010 12:00:00 AM

Department of Systems Engineering and Engineering Management
&
CUHK MoE-Microsoft Key Laboratory of Human-Centric Computing and Interface Technologies

Distinguished Lecture

Title: Crowds and Clouds for Spoken Language Processing
Date: 22 November 2010 (Monday)
Time: 9:30 am - 10:45 am
Venue: Room 513, William M. W. Mong Engineering Building, The Chinese University of Hong Kong
Speaker:
     Professor Jim Glass
     Principal Research Scientist
     MIT Computer Science and Artificial Intelligence Laboratory
     Massachusetts Institute of Technology

Abstract:
In this talk I discuss how crowdsourcing techniques and cloud-based services can lead to a paradigm shift in spoken language processing development.  Through these new opportunities, distributed and unstructured knowledge sources can be leveraged by exploiting weakly-supervised learning in place of expert supervision. I illustrate several different ways in which crowdsourcing can be used to accelerate spoken language technology development, including data collection and annotation, as well as system assessment.  I also present recent work that highlights how crowdsourcing can be used for learning new word pronunciations, and how unstructured opinions can be summarized to enhance spoken information access.  These capabilities will enable a new breed of spoken language technology that can operate semi-autonomously in a cloud-based environment and use crowdsourcing techniques to expand their capabilities.  The talk will include several demonstrations of prototype web-based spoken dialogue systems that leverage an open-source framework we have developed to deploy spoken language technology on a variety of web-enabled platforms.
 

About the speaker:
James Glass is a Principal Research Scientist at the MIT Computer Science and Artificial Intelligence Laboratory where he heads the Spoken Language Systems Group.  He is also a Lecturer in the Harvard-MIT Division of Health Sciences and Technology.  His primary research interests are in the area of speech communication and human-computer interaction, centered on automatic speech recognition and spoken language understanding.
 

 


 

Seminar on Learning Speech Patterns by Professor Jim Glass 11/25/2010 12:00:00 AM 11/25/2010 12:00:00 AM

Seminar

Date: 25 Nov 2010
Time: 10:30 – 11:30 a.m.
Venue: Rm 906, ERB
 
Title: Learning Speech Patterns
Speaker: Prof Jim Glass
 
Abstract:
 
The development of an automatic speech recognizer is typically a highly supervised process involving the specification of phonetic inventories, lexicons, acoustic and language models, along with annotated training corpora.  Although some model parameters may be modified via adaptation, the overall structure of the speech recognizer remains relatively static thereafter.  While this approach has been effective for problems when there is adequate human expertise and labelled corpora, it is challenged by less-supervised or unsupervised scenarios.  It also stands in stark contrast to human processing of speech and language where learning is an intrinsic capability.
 
From a machine learning perspective, a complementary alternative is to discover unit inventories in a less supervised manner by exploiting the structure of repeating acoustic patterns within the speech signal. In this talk I describe a pattern discovery method to automatically acquire lexical entities and speaker segmentations directly from an untranscribed audio stream.  Our approach to unsupervised word acquisition utilizes a segmental variant of a widely used dynamic programming technique, which is used to find matching acoustic patterns between spoken utterances.  By aggregating information about these matching patterns across audio streams, it is possible to group similar acoustic sequences together to form clusters corresponding to lexical entities such as words and short multi-word phrases.  Clusters found by applying this technique on a corpus of academic lectures exhibit high purity; many of the corresponding lexical identities are relevant to the underlying audio stream.  This method has also been applied to languages other than English, and can also be used to perform unsupervised speaker segmentation.
 
Joint work with Alex Park and Yaodong Zhang
 
Bio:
 
James Glass is a Principal Research Scientist at the MIT Computer Science and Artificial Intelligence Laboratory where he heads the Spoken Language Systems Group.  He is also a Lecturer in the
Harvard-MIT Division of Health Sciences and Technology.  His primary research interests are in the area of speech communication and human-computer interaction, centered on automatic speech recognition and spoken language understanding.

 


 

Seminar on "Language Model Combination and Adaptation using Weighted Finite State Transducers" by Dr. Andrew Liu 2/23/2010 12:00:00 AM 2/23/2010 12:00:00 AM

Seminar

Date: 23 February 2010
Time: 4:00
Venue: ERB513

Title: Language Model Combination and Adaptation using Weighted Finite State Transducers
Speaker: Dr. Andrew Liu
                 Research Associate
                 Cambridge University

Abstract:
In speech recognition systems language model (LMs) are often constructed by training and combining multiple n-gram models. They can be either used to represent different genres or tasks found in diverse text sources, or capture stochastic properties of different linguistic symbol sequences, for example, syllables and words. Unsupervised LM adaptation may also be used to further improve robustness to varying styles or tasks. When using these techniques, extensive software changes are often required. In this paper an alternative and more general approach based on weighted finite state transducers (WFSTs) is investigated for LM combination and adaptation. As it is entirely based on well-defined WFST operations, minimum change to decoding tools is needed. A wide range of LM combination configurations can be flexibly supported. An efficient on-the-fly WFST decoding algorithm is also proposed. Significant error rate gains of 7.3% relative were obtained on a state-of-the-art broadcast audio recognition task using a history dependently adapted multi-level LM modelling both syllable and word sequences.

NULL 0 1 110 1 LECTURE 12/25/2010 11:51:53 PM 12/25/2010 11:54:47 PM Distinguished Lecture on We Are Drowning in Multimedia Data! Hurray!! by Dr. Malcolm Slaney 10/20/2010 12:00:00 AM 10/20/2010 12:00:00 AM

Distinguished Lecture

Date: 20 October 2010
Time: 4:30pm
Venue: ERB513

Title: We Are Drowning in Multimedia Data! Hurray!!
Speaker: Dr. Malcolm Slaney
                 Principal Scientist
                 Yahoo! Research

Abstract:
The wealth of data available on the Internet changes the way we think about multimedia. Never before has there been so much multimedia data available for training models and answering questions. But these new riches bring with it a change in the problems we must  think about. The data is noisy and largely unlabeled --- we must make sense of it, often  returning an answer in hundreds of milliseconds. How do we understand the user's environment, especially when it extends across the world?  How do we take into account context  and do it at the scale of the Internet? In this talk I'd like to share with you Yahoo's experiences in this brave new world of multimedia everywhere, describe promising new technologies, and discuss open research directions. I will describe the need for better user and multimedia models, the kinds of algorithms needed for today's large databases, and how the Internet is changing multimedia retrieval.

Biography:
Malcolm Slaney is a principal scientist at Yahoo! Research Laboratory. He received his PhD from Purdue University for his work on computed imaging. He is a coauthor, with A. C. Kak, of the IEEE ook "Principles of Computerized Tomographic Imaging." This book was
recently republished by SIAM in their "Classics in Applied Mathematics" Series. He is coeditor, with Steven Greenberg, of the book "Computational Models of Auditory Function."  Before Yahoo!, Dr. Slaney has worked at Bell Laboratory,  Schlumberger Palo Alto Research, Apple Computer, Interval Research  and IBM's Almaden Research Center. He is also a (consulting) Professor at Stanford's CCRMA where he organizes and teaches the Hearing Seminar. His research interests include auditory modeling and perception,  multimedia analysis and synthesis, compressed-domain processing, music similarity and audio search, and machine learning. For the last several years he has lead the auditory group at the Telluride Neuromorphic Workshop

 

 


 

Round Table Meeting with MSRA Managing Director Dr. Hsiao-Wuen Hon 4/7/2011 12:00:00 AM 4/7/2011 12:00:00 AM

Round Table Meeting

Date: 7 April 2011 (Thursday)
Time: 2:00pm
Venue: Room 222, Ho Sin Hang Engineering Building, CUHK
Speaker: Dr. Hsiao-Wuen Hon

Content: know more about research and internship opportunities at MSRA

NULL 0 1 112 1 FORUM 6/2/2011 5:09:18 PM 6/2/2011 5:09:18 PM The Sixth Beijing-Hong Kong International Doctoral Forum 2011 8/1/2011 12:00:00 AM 8/4/2011 12:00:00 AM

The 6th Beijing-Hong Kong International Doctoral Forum 2011

~Clouding Computing, Web, Network and Rich Media Computing~

1-4 August 2011
Shenzhen and Hong Kong, China

The 6th Beijing-Hong Kong International Doctoral Forum 2011 invites you to experience both the vibrant and cosmopolitan atmosphere of Shenzhen and Hong Kong while enjoying the high quality forum. The objective is to bring together Ph.D. students from all over the world, contributing to all facets of the web, network and rich media computing etc. Expenses within the forum for the attendees will be covered by GraduateSchool at Shenzhen, TsinghuaUniversity and The Chinese University of Hong Kong. This includes registration, meals, accommodation, local transport, etc.

Important Dates
Paper submission deadline       June 20, 2011
Notification of acceptance         July 5, 2011
Registration                                 July 12, 2011
Camera-ready paper due date July 17, 2011
Forum                                          August 1-4, 2011

 

 

Please visit http://wiki.cse.cuhk.edu.hk/PHDForum/2011/start for more information

Date: 
Tuesday, October 25, 2011 - 14:30