ASR2K: Speech Recognition for Around 2000 Languages without Audio.
Xinjian LiFlorian MetzeDavid R. MortensenAlan W. BlackShinji WatanabePublished in: CoRR (2022)
Keyphrases
- speech recognition
- speaker identification
- automatic speech recognition
- language identification
- speech processing
- speech recognition technology
- broadcast news
- speech signal
- language model
- hidden markov models
- noisy environments
- speech recognizer
- pattern recognition
- speech synthesis
- cepstral coefficients
- acoustic features
- speech retrieval
- audio visual speech recognition
- word error rate
- multimedia
- handwriting recognition
- speech recognizers
- speech recognition systems
- voice activity detection
- english text
- speaker recognition
- speaker independent
- audio signal
- speaker dependent
- signal processing
- language independent
- visual information
- speaker diarization
- spontaneous speech
- conversational speech
- natural language
- neural network