Exploring Cross-Utterance Speech Contexts for Conformer-Transducer Speech Recognition Systems

作者：Mingyu Cui, Mengzhe Geng, Jiajun Deng, Chengji Deng, Jiawen Kang, Shujie Hu, Guinan Li, Tianzi Wang, Zhaoqing Li, Xie Chen, Xunying Liu · 发表于：IEEE Transactions on Audio Speech and Language Processing · 年份：2025 · DOI：10.1109/taslpro.2025.3606235 · 被引用次数：2 · 研究领域：Speech Recognition and Synthesis

This paper investigates four types of cross-utterance speech contexts modeling approaches for streaming and non-streaming Conformer-Transformer (C-T) ASR systems: i) input audio feature concatenation; ii) cross-utterance Encoder embeddings concatenation; iii) cross-utterance Encoder embeddings pooling projection; or iv) a novel chunk-based approach applied to C-T models for the first time. An efficient batch training scheme is proposed for contextual C-Ts that uses spliced speech utterances within each minibatch to minimize the synchronization overhead while preserving the sequential order of cross-utterance speech contexts. Experiments are conducted on four benchmark speech datasets across three languages: the English GigaSpeech and Mandarin Wenetspeech corpora used in contextual C-T models pre-training; and the English DementiaBank Pitt and Cantonese JCCOCC MoCA elderly speech datasets used in domain fine-tuning. The best performing contextual C-T systems consistently outperform their respective baselines using no cross-utterance speech contexts in pre-training and fine-tuning stages with statistically significant average word error rate (WER) or character error rate (CER) reductions up to0.9%,1.1%,0.51%, and0.98%absolute (6.0%,5.4%,2.0%, and3.4%relative) on the four tasks respectively. Their performance competitiveness against Wav2vec2.0-Conformer, XLSR-128, and Whisper models highlights the potential benefit of incorporating cross-utterance speech contexts into current sp...