Zipformer: A faster and better encoder for automatic speech recognition.
Zengwei YaoLiyong GuoXiaoyu YangWei KangFangjun KuangYifan YangZengrui JinLong LinDaniel PoveyPublished in: CoRR (2023)
Keyphrases
- automatic speech recognition
- speech recognition
- word error rate
- hidden markov models
- speech signal
- conversational speech
- speech retrieval
- recognition errors
- spontaneous speech
- noisy environments
- broadcast news
- bit rate
- spoken words
- speech corpus
- word recognition
- acoustic features
- discriminative training
- neural network
- language model
- multiresolution
- natural language
- pattern recognition
- computer vision
- information retrieval
- speaker adaptation
- machine learning