Exploring Capabilities of Monolingual Audio Transformers using Large Datasets in Automatic Speech Recognition of Czech.
Jan LeheckaJan SvecAles PrazákJosef PsutkaPublished in: INTERSPEECH (2022)
Keyphrases
- automatic speech recognition
- cl sr
- speech retrieval
- broadcast news
- speech recognition
- cross language
- spoken document retrieval
- conversational speech
- spontaneous speech
- speech signal
- acoustic features
- document collections
- manually generated
- noisy environments
- hidden markov models
- multimedia
- speaker identification
- word error rate
- question answering
- recognition errors
- spoken words
- word recognition
- cross lingual
- music information retrieval
- visual information
- query expansion