The Fosafer System for The ICASSP2024 In-Car Multi-Channel Automatic Speech Recognition Challenge
Shangkun Huang, Yuxuan Du, Yankai Wang, Jing Deng, Rong Zheng · 2024
This paper presents the Fosafer’s submissions to the ICASSP 2024 In-Car Multi-Channel Automatic Speech Recognition Challenge (ICMC-ASR), which includes both the Automatic Speech Recognition (ASR) and Automatic Speech Diarization and Recognition (ASDR) systems. In Track1, a robust ASR system with data augmentation, self-supervised learning representation (SSLR), and speech enhancement (SE) achieved the second place. In Track2, different speaker diarization algorithms were fully exploited and achieved the fifth place.