通过筛选和选择弱监督数据，显著降低语音识别错误率。
Improving Large-Scale Weakly Supervised ASR by Filtering and Selection
- 先全集预训练，再按字符错误率筛选子集继续预训练
- 在目标领域相似样本中精选微调，使错误率下降6.4%和4.0%
- 适合大规模弱监督语音识别任务，尤其对低资源语言有效
利用大规模弱监督数据集对于训练鲁棒的端到端自动语音识别（ASR）模型至关重要。然而，这类数据集常含噪声标签且缺乏领域特异性，限制了其有效性。为解决这些问题并更好利用弱监督数据，我们提出一种结合数据过滤与选择的新训练方法。该方法包含三步：在全量数据上预训练，基于字符错误率（CER）在筛选后的子集上继续预训练，并从筛选子集中选出与目标领域声学特性相近的小样本进行微调。在9万小时的弱监督日语数据集上的实验表明，所提出的过滤与选择方法分别协同降低了6.4%和4.0%的CER，尽管这些步骤复用了首次预训练中已使用的样本。
原文摘要 · Abstract (English)
Leveraging large-scale weakly supervised datasets is crucial to train robust end-to-end automatic speech recognition (ASR) models. However, such datasets often contain noisy labels and lack domain specificity, limiting their effectiveness. To address these issues and make better use of weakly supervised datasets, we propose a novel training approach incorporating data filtering and selection. Our approach consists of three steps: pretraining on the entire dataset, continued pretraining on a filtered subset based on character error rate (CER), and fine-tuning on a small number of acoustically similar samples to the target domain, selected from the filtered subset. In experiments with a 90,000-hour weakly supervised Japanese dataset, the proposed filtering and selection methods synergistically reduced CER by up to 6.4% and 4.0%, respectively, even though these steps reused training samples already used in the first pretraining step.
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。