The USTC-Nercslip Systems for the ICMC-ASR Challenge

• Published in: 2024 IEEE International Conference on Acoustics, Speech, and Signal Processing Workshops (ICASSPW) pp. 3 - 4
• Main Authors: Wu, Minghui, Xu, Luzhen, Zhang, Jie, Tang, Haitao, Yue, Yanyan, Liao, Ruizhi, Zhao, Jintao, Zhang, Zhengzhe, Wang, Yichi, Yan, Haoyin, Yu, Hongliang, Ma, Tongle, Liu, Jiachen, Wu, Chongliang, Li, Yongchao, Zhang, Yanyong, Fang, Xin, Zhang, Yue
• Format: Conference Proceeding
• Language: English
• Published: IEEE 14.04.2024
• Subjects: Acoustics; Array signal processing; Conferences; Data models; ICMC-ASR challenge; Iterative methods; Linguistics; multi-channel beamforming; pseudo-label; Self-supervised learning; speaker diarization
• DOI: 10.1109/ICASSPW62465.2024.10627164

This report describes the submitted system to the In-Car Multi-Channel Automatic Speech Recognition (ICMC-ASR) challenge, which considers the ASR task with multi-speaker overlapping and Mandarin accent dynamics in the ICMC case. We implement the front-end speaker diarization using the self-supervised learning representation based multi-speaker embedding and beamforming using the speaker position, respectively. For ASR, we employ an iterative pseudo-label generation method based on fusion model to obtain text labels of unsupervised data. To mitigate the impact of accent, an Accent-ASR framework is proposed, which captures pronunciation-related accent features at a fine-grained level and linguistic information at a coarse-grained level. On the ICMC-ASR eval set, the proposed system achieves a CER of 13.16% on track 1 and a cpCER of 21.48% on track 2, which significantly outperforms the official baseline system and obtains the first rank on both tracks.