Training automatic speech recognition (ASR) systems requires large amounts of well-curated paired data. However, human annotators usually perform "non-verbatim" transcription, which can result in poorly trained models. In this paper, we propose Omni-temporal Classification (OTC), a novel training criterion that explicitly incorporates label uncertainties originating from such weak supervision. This allows the model to effectively learn speech-text alignments while accommodating errors present in the training transcripts. OTC extends the conventional CTC objective for imperfect transcripts by leveraging weighted finite state transducers. Through experiments conducted on the LibriSpeech and LibriVox datasets, we demonstrate that training ASR models with OTC avoids performance degradation even with transcripts containing up to 70% errors, a scenario where CTC models fail completely. Our implementation is available at https://github.com/k2-fsa/icefall.

翻译：训练自动语音识别系统需要大量精心整理的对齐数据。然而，人工标注者通常执行"非逐字"转录，这可能导致模型训练效果不佳。本文提出了一种全新的训练准则——全时态分类（OTC），该准则明确地将此类弱监督产生的标签不确定性纳入考量。这使得模型能够在容纳训练转录文本中存在的错误的同时，有效学习语音-文本对齐关系。OTC通过利用加权有限状态换能器，扩展了针对不完美转录文本的传统CTC目标函数。通过在LibriSpeech和LibriVox数据集上进行的实验，我们证明了即使转录文本包含高达70%的错误（在此场景下CTC模型完全失效），使用OTC训练的ASR模型也能避免性能下降。我们的实现代码可访问https://github.com/k2-fsa/icefall获取。