针对噪声标注数据，提出三类新损失函数提升语音识别鲁棒性。
RNN-Transducer-based Losses for Speech Recognition on Noisy Targets
- 引入跳帧与跳词机制，在损失图中修复误删和误插错误。
- 在噪声数据上恢复超过70%的识别性能，接近准确标注水平。
- 适合工业级大规模语音识别训练，尤其标注质量参差的场景。
在工业级语音识别系统中，由于数据量庞大，难以确保每条语料都拥有精确转录，因此在噪声标注数据上训练面临巨大挑战。本文针对RNN-Transducer模型提出新型损失函数，以减轻转录错误的影响。星型-转换器损失（Star-Transducer）通过在损失图中引入“跳帧”转移，有效修复删除错误，使系统性能恢复至准确标注模型的90%以上。旁路-转换器损失（Bypass-Transducer）采用“跳词”转移应对插入错误，恢复超过60%的识别质量。最终提出的目标鲁棒转换器损失（Target-Robust Transducer）融合两种机制，对任意类型的标注错误均表现出强鲁棒性。实验表明，该损失函数在噪声数据上使RNN-T性能恢复超过70%，显著优于基线方法。
原文摘要 · Abstract (English)
Training speech recognition systems on noisy transcripts is a significant challenge in industrial pipelines, where datasets are enormous and ensuring accurate transcription for every instance is difficult. In this work, we introduce novel loss functions to mitigate the impact of transcription errors in RNN-Transducer models. Our Star-Transducer loss addresses deletion errors by incorporating "skip frame" transitions in the loss lattice, restoring over 90% of the system's performance compared to models trained with accurate transcripts. The Bypass-Transducer loss uses "skip token" transitions to tackle insertion errors, recovering more than 60% of the quality. Finally, the Target-Robust Transducer loss merges these approaches, offering robust performance against arbitrary errors. Experimental results demonstrate that the Target-Robust Transducer loss significantly improves RNN-T performance on noisy data by restoring over 70% of the quality compared to well-transcribed data.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。