An Inexpensive Packet Loss Compensation Scheme for Distributed Speech Recognition Based on Soft-Features.
Valentin IonReinhold Haeb-UmbachPublished in: ICASSP (1) (2006)
Keyphrases
- speech recognition
- packet loss
- speech recognition systems
- pattern recognition
- hidden markov models
- feature set
- error propagation
- packet loss rate
- language model
- speech signal
- video transmission
- speech synthesis
- feature vectors
- video quality
- multiple description coding
- automatic speech recognition
- error control
- noisy environments
- feature space
- congestion control
- speech recognizer
- network congestion
- feature extraction
- image data
- wireless channels
- error resilience
- speaker identification
- information retrieval
- video sequences