Core Research - Multimedia Signal Processing

Multimedia Signal Processing
Projects/Technologies Audio Beam System: The next generation (v 2.0) - PI: A/P Gan Woon Seng (ewsgan@ntu.edu.sg) Diagram shows the first generation of the Audio Beam System - Previous work on the development of a directional sound beam: Audio Beam System IES Prestigious Engineering Achievement Award 2001 Publicity in Straits Time, Channel News Asia and Chinese media More than 10 journal publications, and 4 local and US patents granted - Currently, research and develop on the Next Generation of Directional Sound Beam with Bass Enhancement and Beamsteering to Support New Interactive Digital Media Applications - Awarded the First Interactive & Digital Media (IDM) Research and Development Programme Grant from National Research Foundation (NRF)
Speech Touch and Acoustic Tangible Interfaces for Next-generation Applications (STATINA) - Ast./P Andy Khong, A/P Gan Woon Seng
Project Resources: Funding from NRF-IDM (Co-Space) 2 RFs, 3 RAs/POs, 2 Ph.D. start date: 15th December 2008 for 3 years an excellent platform for collaborating with A*Star and UIUC through co-superivision of Ph.Ds Key Significance: Conversion of daily objects (tabletops, glass panels) into touch panels using discrete surface mounted sensors Project Objectives: Investigation of in-solid acoustic propagation for touch technology Speech recognition for system wake-up Audio beam technology for audio feedback Audio enhancement for enhanced quality Impact and Applications: Human-machine interfaces for mobile/stationed devices Rehabilitation for the disabilities Robotic control Instant messenging (IM) applications Collaborators: Georgia Tech. Northern Illinois Univ.
Multimodal Video Content Analysis, Modeling and Content Creation for Mobile Devices - A/P Xue Ping and Dr. Tian Qi(I2R) Funding Source: A*STAR, Science and Engineering Research Council (SERC) Objective: Investigate and develop algorithms, tools and systems for multimodal video (with accompanied audio) content analysis, modeling and creation for mobile devices Video content analysis - low level features and semantic features Video content creation tools – modeling and personalization Prototype systems
Create video summary with index to events found in a video. Inputs: Video file & Query: The user picks the event type to be seen in the summarized video. Output: The summarized video with corresponding event types.	Detect and remove audio ads and replace the detected segment with targeted ads. Inputs: Various compressed audio formats such as mp3, AAC. etc. Output: with targeted ads which may contain a replaced ad and/or an inserted ad according to user’s interests.