arXiv cs.CL· Fu-An Chao, Bi-Cheng Yan, Berlin Chen·· 6 天前AI 评分22
OpenPA：无需强制对齐的混合转写自适应多粒度发音评估框架
OpenPA: Forced-alignment-free Multi-granular Pronunciation Assessment with Mixed-transcript Adaptation
AI 导读
OpenPA 是一个无需强制对齐的自动发音评估框架，利用 Whisper 的交叉注意力机制同时预测词级和语句级发音分数。它通过混合转写自适应和自适应替换加权，在单一模型中兼容封闭与开放响应场景的异构转写。仅用 speechocean762 训练，OpenPA 在域内封闭及模拟开放响应条件下取得所评估系统中最强整体表现，并在 MultiPA 上保持零样本性能。
正文 · 原文
Abstract:Recent advances in automatic pronunciation assessment (APA) have driven a shift from scripted to free-form assessment. However, existing methods often rely on multi-stage pipelines with forced alignment and separate feature extractors, limiting their practicality across response settings. In light of this, we propose OpenPA, a novel APA framework that leverages Whisper's cross-attention mechanism to jointly predict word- and utterance-level pronunciation scores without forced alignment. Through mixed-transcript adaptation and adaptive substitution weighting, OpenPA accommodates heterogeneous transcripts across closed- and open-response settings within a single model. Trained solely on speechocean762, OpenPA achieves the strongest overall performance among evaluated systems under in-domain closed and simulated open-response conditions, while maintaining its performance zero-shot on MultiPA. An additional cross-task evaluation on oral-cancer speech suggests its potential applicability beyond non-native speech.
| Comments: | submitted to ICASSP 2027 |
| Subjects: | Audio and Speech Processing (eess.AS); Computation and Language (cs.CL); Sound (cs.SD); Signal Processing (eess.SP) |
| Cite as: | arXiv:2610.13826 [eess.AS] |
| (or arXiv:2610.13826v1 [eess.AS] for this version) | |
| https://doi.org/10.48550/arXiv.2610.13826 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Fu-An Chao [view email]
[v1]
Fri, 9 Oct 2026 16:35:25 UTC (1,909 KB)
来源：arXiv cs.CL · arxiv.org