arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.16664cs.CLcs.SDeess.AS
收藏

分数平衡损失用于多维度发音评估

Score-balanced Loss for Multi-aspect Pronunciation Assessment

Graduate School of AI, POSTECH(浦项科技大学人工智能研究生院);Department of Computer Science and Engineering, POSTECH(浦项科技大学计算机科学与工程系)

Heejin Do, Yunsu Kim, Gary Geunbae Lee

AI总结:

针对发音评估中分数标签不平衡问题,提出分数平衡损失函数,通过有效样本数和分数排名设计加权因子,在speechocean762数据集上验证了有效性。

摘要:

随着技术的快速发展,自动发音评估已转向从多个方面(如流利度和重音)评估发音的系统。然而,尽管每个方面内的分数标签高度不平衡,现有研究很少解决数据不平衡问题。在本文中,我们提出了一种新颖的损失函数,即分数平衡损失,以解决由不均匀数据引起的问题,例如对多数分数的偏向。作为一种重新加权方法,当预测分数属于少数类时,我们分配更高的成本,从而引导模型在稀疏分数预测中获得正反馈。具体而言,我们利用有效样本数的概念和分数排名来设计两个加权因子。我们在speechocean762数据集上评估了我们的方法,该数据集在几个方面具有明显不平衡的分数。特别是在这些不均匀方面上的改进结果证明了我们方法的有效性。

英文摘要:

With rapid technological growth, automatic pronunciation assessment has transitioned toward systems that evaluate pronunciation in various aspects, such as fluency and stress. However, despite the highly imbalanced score labels within each aspect, existing studies have rarely tackled the data imbalance problem. In this paper, we suggest a novel loss function, score-balanced loss, to address the problem caused by uneven data, such as bias toward the majority scores. As a re-weighting approach, we assign higher costs when the predicted score is of the minority class, thus, guiding the model to gain positive feedback for sparse score prediction. Specifically, we design two weighting factors by leveraging the concept of an effective number of samples and using the ranks of scores. We evaluate our method on the speechocean762 dataset, which has noticeably imbalanced scores for several aspects. Improved results particularly on such uneven aspects prove the effectiveness of our method.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑