|
Senior Researcher Qifeng Zhu received the B.S. degree in Tsinghua University, Beijing, China in 1994, M.S in the Institute of Acoustics of Chinese Academy of Science in 1997, and Ph.D. in University of California, Los Angeles in 2001. Currently he is working on the Novel Approaches of the EARS (effective affordable, reusable speech-to-text) project sponsored by DARPA (defense advanced research projects agency) of DoD. He worked at speech R&D of Nuance Communications as a consultant and a research engineer in the speech R&D from 2000 to 2002. |
|
Research interests Publications CV Free Software Personal web at UCLA
Dr. Zhu’s research interests focus on speech recognition and understanding related issues, including speech signal processing and feature extraction, noise robustness, and speech modeling. Some current research directions include embedded speech recognition system, incorporating knowledge in ASR system, and issues on system generalization by modeling fundamentals in speech.
His long-term research interests include general pattern analysis and recognition problem in computer vision, bio-engineering, and bio-metrics, and intelligent systems which include components of perception/understanding/decision/reaction.
An important accomplishment in his Ph.D. research at UCLA is on novel algorithms on noise robust feature extraction. The performance is evaluated in the Aurora 2 Evaluation at Eurospeech2001. In the recent two years he is working on continuous telephone speech recognition, and noise in cars at Nuance Communications. Currently he is working on novel features other than the traditional energy envelope based features for speech recognition in the EARS project, at ICSI, affiliated with UC Berkeley.
My personal web is at http://www.ee.ucla.edu/~qifeng
Updated on 2/2003