针对6种南部班图语语音识别难题，提出基于语调的渐进式训练方法。
Tone-Conditioned Curriculum Learning for Low-Resource Bantu Speech Recognition
- 根据语调统计动态调整训练难度，结合门控适配器优化模型
- 在跨数据集测试中平均词错误率降至28.41%，西茨onga语达23.79%
- 不同语言需匹配特定模型，建议按语种选型并多数据集验证
南部班图语有超过8000万人使用，但现有基础语音识别模型在零样本条件下词错误率（WER）仍高于100%，难以应用于教育与公共服务。本文针对6种南部班图语，提出一种基于语调条件的渐进式学习框架，融合混合难度评分、由语调统计驱动的门控适配器及分阶段课程训练。在社区语料库上训练，并在NCHLT数据集上测试跨域泛化能力。结果表明，架构与语言间存在明显交互：在宁加语系语言上，W2V-BERT比Whisper低3至4个WER点；而在索托-茨瓦纳语系上Whisper表现更优。采用语调条件的W2V-BERT在所有数据集上平均达到28.41% WER，Xitsonga迁移测试中为23.79%。无单一模型适用于全部6种语言，部署时应按语言选择模型，并在多语料上验证效果。
原文摘要 · Abstract (English)
Southern Bantu languages are spoken by over 80 million people, yet current foundation ASR models still produce zero-shot WER above 100%, which limits practical use in education and public services. We addressed this gap with a tone conditioned curriculum framework for 6 Southern Bantu languages that combined hybrid difficulty scoring, gated adapters driven by tonal statistics and staged curriculum training. We trained on a community corpus and tested transfer to NCHLT to measure robustness beyond matched evaluation. Results revealed clear interactions between architecture and language, with W2V-BERT outperforming Whisper on Nguni languages by 3 to 4 WER points whilst Whisper performed better on Sotho-Tswana languages. W2V-BERT with tone conditioning reached 28.41% average WER across datasets and 23.79% on Xitsonga transfer. No single model suited all 6 languages, so deployment should pair model selection per language with validation across corpora.
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。