Unified probabilistic approach to error concealment for distributed speech recognition.
Valentin IonReinhold Haeb-UmbachPublished in: INTERSPEECH (2005)
Keyphrases
- speech recognition
- error concealment
- language model
- video transmission
- visual quality
- hidden markov models
- pattern recognition
- speech recognizer
- automatic speech recognition
- speech signal
- temporal error concealment
- speech synthesis
- packet loss
- motion vectors
- video quality
- probabilistic model
- speaker identification
- speaker independent
- speech recognition technology
- bayesian networks
- bitstream
- speech recognition systems
- neural network
- video compression
- human visual system
- image quality