Robustness in recognizing human speech by machines has been one of the most challenging research problems in Automatic Speech Recognition (ASR) for the past several decades. Significant efforts have been directed towards making ASR technology more robust to variations due to regional dialects, background noise, and input device and transmission channel characteristics, as well as variability among speakers due to age, intonation and co-articulation effects. Yet, even with these efforts, ASR technology is still relatively fragile to the above variations, making voice-enabled systems difficult to deploy and, therefore, impractical to scale. What is missing from our research that is holding these deployments back? This panel will attempt to answer this question from the perspective of those with experience in deploying ASR technologies.
A satellite workshop to ICASSP 2005 is currently being planned to expand on this topic. Further details will be announced at the panel session.
Mazin Rahim, AT&T Labs-Research
180 Park Avenue, Florham Park, NJ 07932, USA
Tel: +1 973-360-8529
Larry Heck, Nuance Communication
1005 Hamilton Ct
Menlo Park, CA 94025
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Paper Submission -||- Special Sessions -||- ITT -||- Paper Review -||- Exhibits -||- Tutorials
Information -||- Registration -||- Travel Insurance -||- Housing -||- Workshops