Guided contrastive self-supervised pre-training for automatic speech recognition.
Aparna KhareMinhua WuSaurabhchand BhatiJasha DroppoRoland MaasPublished in: CoRR (2022)
Keyphrases
- automatic speech recognition
- discriminative training
- speech recognition
- hidden markov models
- speech signal
- broadcast news
- speech retrieval
- conversational speech
- acoustic models
- spoken words
- word error rate
- spontaneous speech
- recognition errors
- acoustic features
- noisy environments
- speech corpus
- training process
- word recognition
- non stationary
- computer vision