WST: Weakly Supervised Transducer for Automatic Speech Recognition
This addresses the costly data annotation bottleneck for ASR systems, offering a robust solution for industrial applications.
The paper tackles the problem of reducing reliance on high-quality annotated data in automatic speech recognition by proposing a Weakly Supervised Transducer (WST), which maintains performance with transcription error rates up to 70% and outperforms existing weakly supervised methods like BTC and OTC.
The Recurrent Neural Network-Transducer (RNN-T) is widely adopted in end-to-end (E2E) automatic speech recognition (ASR) tasks but depends heavily on large-scale, high-quality annotated data, which are often costly and difficult to obtain. To mitigate this reliance, we propose a Weakly Supervised Transducer (WST), which integrates a flexible training graph designed to robustly handle errors in the transcripts without requiring additional confidence estimation or auxiliary pre-trained models. Empirical evaluations on synthetic and industrial datasets reveal that WST effectively maintains performance even with transcription error rates of up to 70%, consistently outperforming existing Connectionist Temporal Classification (CTC)-based weakly supervised approaches, such as Bypass Temporal Classification (BTC) and Omni-Temporal Classification (OTC). These results demonstrate the practical utility and robustness of WST in realistic ASR settings. The implementation will be publicly available.