Integration of Frame- and Label-synchronous Beam Search for Streaming Encoder-decoder Speech Recognition.
Emiru TsunooHayato FutamiYosuke KashiwagiSiddhant AroraShinji WatanabePublished in: INTERSPEECH (2023)
Keyphrases
- speech recognition
- beam search
- wyner ziv
- distributed video coding
- pixel domain
- video codec
- bit budget
- rate distortion
- hidden markov models
- heuristic search
- search algorithm
- bit rate
- speech synthesis
- branch and bound
- language model
- video coding
- bitstream
- search problems
- search methods
- automatic speech recognition
- speech recognizer
- speech signal
- hill climbing
- speaker identification
- speech recognition systems
- video quality
- ranking functions
- motion vectors
- pattern recognition
- compressed domain
- search space
- error concealment
- information retrieval
- document retrieval
- motion estimation
- feature selection
- computer vision