对比Whisper与Wav2Vec-BERT在孟加拉语上的表现，发现后者更高效准确。
Adaptability of ASR Models on Low-Resource Language: A Comparative Study of Whisper and Wav2Vec-BERT on Bangla
- 用公共数据集微调两个模型，优化学习率等超参数。
- Wav2Vec-BERT的词错误率和字符错误率均更低，训练更快。
- 适合资源匮乏语言的语音识别系统开发者参考。
近年来，基于大规模多语言文本和语音数据训练的神经模型在支持低资源语言方面展现出巨大潜力。本研究考察了两种前沿自动语音识别（ASR）模型——OpenAI的Whisper（Small & Large-V2）和Facebook的Wav2Vec-BERT在孟加拉语这一低资源语言上的表现。实验使用Mozilla Common Voice-17和OpenSLR两个公开数据集，通过系统的微调与超参数优化（包括学习率、训练轮数及模型检查点选择），从词错误率（WER）、字符错误率（CER）、训练时间和计算效率四个方面进行对比。结果表明，Wav2Vec-BERT在所有关键评估指标上均优于Whisper，且所需计算资源更少，为低资源语言环境下构建稳健语音识别系统提供了重要参考。
原文摘要 · Abstract (English)
In recent years, neural models trained on large multilingual text and speech datasets have shown great potential for supporting low-resource languages. This study investigates the performances of two state-of-the-art Automatic Speech Recognition (ASR) models, OpenAI's Whisper (Small & Large-V2) and Facebook's Wav2Vec-BERT on Bangla, a low-resource language. We have conducted experiments using two publicly available datasets: Mozilla Common Voice-17 and OpenSLR to evaluate model performances. Through systematic fine-tuning and hyperparameter optimization, including learning rate, epochs, and model checkpoint selection, we have compared the models based on Word Error Rate (WER), Character Error Rate (CER), Training Time, and Computational Efficiency. The Wav2Vec-BERT model outperformed Whisper across all key evaluation metrics, demonstrated superior performance while requiring fewer computational resources, and offered valuable insights to develop robust speech recognition systems in low-resource linguistic settings.
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。