Conference XP Project
Sound source localization by microphone arrays
Automatic calibration of microphone arrays
Audio-visual tracking of speakers
Tele-conferencing system with automatic focusing using microphone arrays

Web Object Indexing Using Domain Knowledge
Web object is defined to represent any meaningful object embedded in web pages (e.g. images, music) or pointed to by hyperlinks (e.g. downloadable files). Users usually search for information of a certain ‘object’, rather than a web page containing the query terms. To facilitate web object searching and organizing, we propose a novel approach to web object indexing, by detecting its inherent structure information with domain knowledge. In our approach, Layered LSI (Latent Semantic Indexing) spaces are built for the hierarchical-structured domain knowledge, in order to emphasize specific semantics and term space in each layer of domain knowledge. Web object representation is constructed by text and hyperlink analysis; and then further pruned to remove the noises. The structure attributes of web object are finally extracted by finding the knowledge document that best matches the web object. Our approach also indicates a new way to use trust-worthy Deep Web knowledge to help organize dispersive information of Surface Web.

Music Repeating Pattern Discovery and Music Structure Analysis
Music and songs usually have repeating patterns and repetitive structure. The automatic extraction of such repeating patterns and structure is useful for further music summarization, indexing and retrieval. We develop an effective approach of repeating pattern discovery and structure analysis is proposed for acoustic music data. Constant Q transform is utilized for feature extraction and a novel similarity measure between musical features is proposed. From the self-similarity matrix of the music, an adaptive method is then presented to extract all the significant repeating patterns. Based on the obtained repetitions, musical structure is further analyzed with some heuristic rules. And finally, an optimization-based approach is proposed to determine the accurate boundary of each musical section.

Music Emotion Recognition
The automatic emotion recognition of music is useful for music understanding, indexing, and retrieval. We present a scheme to recognize the emotional meaning of western tonal music. First, novel musical features are extracted from MIDI files, according to some psychological theories in western cultures. Second, a 2-dimensional ‘Emotion Plane’ model is constructed as music emotion’s taxonomy. Third, the support vector machine (SVM) is trained using the extracted features. Finally, the trained support vector machine can be used to classify the music emotions into six classes, according to the emotion plane model.

A Portable Scene-Aware View Management System for Video-Conferencing
Muyuan WANG, Klara Nahrstedt, Miles Johnson
Accpeted by 2007 8th IEEE International Conference on Multimedia and Expo (ICME'07), Beijing, China, July 2-4,  2007.  
[PDF]

Web Object Indexing Using Domain Knowledge
Muyuan WANG, Zhiwei Li, Lie Lu, Weiying Ma, and Naiyao Zhang
11th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD'05), Chicago, IL USA, August 21-24,  2005.  
[PDF]

Repeating Pattern Discovery and Structure Analysis from Acoustic Music Data
Lie Lu, Muyuan WANG, and Hong-Jiang Zhang
Proc. of 6th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR'04), in conjuction with ACM Multimedia 2004, New York, NY USA, October 15-16,  2004.  
[PDF]

Repeating Pattern Discovery from Acoustic Musical Signals
Muyuan WANG, Lie Lu, and Hong-Jiang Zhang
Proc. of 2004 5th IEEE International Conference on Multimedia and Expo (ICME'04), Taipei, Taiwan, June 27-30, 2004 (Oral Presentation)
[PDF]

User-Adaptive Music Emotion Recognition
WANG Muyuan, ZHANG Naiyao, and ZHU Hancheng
Proc. of 2004 7th International Conference on Signal Processing (ICSP'04), Beijing, China, August 31-September 4, 2004 (Oral Presentation)
[PDF]

Real-time Animation of CAD-MF System Using Simplified Particle System and Fast Translation Algorithm
Mu-Yuan Wang, Nai-yao Zhang, and Xin-yu Du
Proc. of The First International Conference on Machine Learning and Cybernetics (ICMLC'02), Pages:1614-1618 vol.3, Beijing,China, 4-5 November 2002 (Oral Presentation)

Emotion Recognition of Western Tonal Music (In Chinese), Bachelor Thesis, Tsinghua University, June 2003