Hypothesis Stitcher for End-to-End Speaker-attributed ASR on Long-form Multi-talker Recordings
원문 초록
An end-to-end (E2E) speaker-attributed automatic speech recognition (SA-ASR) model was proposed recently to jointly perform speaker counting, speech recognition and speaker identification. The model achieved a low speaker-attributed word error rate (SA-WER) for monaural overlapped speech comprising an unknown number of speakers. However, the E2E modeling approach is susceptible to the mismatch between the training and testing conditions. It has yet to be investigated whether the E2E SA-ASR model works well for recordings that are much longer than samples seen during training. In this work, we first apply a known decoding technique that was developed to perform single-speaker ASR for long-form audio to our E2E SA-ASR task. Then, we propose a novel method using a sequence-to-sequence model, called hypothesis stitcher. The model takes multiple hypotheses obtained from short audio segments that are extracted from the original long-form input, and it then outputs a fused single hypothesis. We propose several architectural variations of the hypothesis stitcher model and compare them with the conventional decoding methods. Experiments using LibriSpeech and LibriCSS corpora show that the proposed method significantly improves SA-WER especially for long-form multi-talker recordings.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Addressing Pitfalls in Auditing Practices of Automatic Speech Recognition Technologies: A Case Study of People with Aphasia
Unsupervised Rhythm and Voice Conversion to Improve ASR on Dysarthric Speech
Towards Temporally Explainable Dysarthric Speech Clarity Assessment
Vedavani: A Benchmark Corpus for ASR on Vedic Sanskrit Poetry
BeaverTalk: Oregon State University's IWSLT 2025 Simultaneous Speech Translation System
Exploring Generative Error Correction for Dysarthric Speech Recognition
토론
이 논문과 연결된 포럼 글이 아직 없습니다.