Variable Frame Rate Acoustic Models Using Minimum Error Reinforcement Learning.
Dongcheng JiangChao ZhangPhilip C. WoodlandPublished in: Interspeech (2021)
Keyphrases
- frame rate
- minimum error
- reinforcement learning
- acoustic models
- video sequences
- high speed
- speech recognition
- bayes optimal
- classification error
- d scene
- bitstream
- automatic speech recognition
- machine learning
- learning process
- hidden markov models
- pattern recognition
- real time
- hyperplane
- learning algorithm
- pairwise
- coding scheme
- computer vision
- speech recognizer