提出一种无需生成的实时标点恢复方法，提升语音识别可读性。
Efficient Punctuation Restoration via Weighted Lookahead Scoring Method for Streaming ASR Systems
- 基于有限前瞻词元的非自回归打分机制，逐字边界决策
- 无微调下达到0.893宏F1，微调后达0.937，优于基线
- 适合低延迟流式语音识别系统，无需推理时参数更新
标点恢复能提升自动语音识别（ASR）结果的可读性。然而流式ASR需在有限未来上下文条件下进行在线决策。现有生成式方法在逐边界评估下易出现延迟高和对齐失败问题。本文提出一种非自回归打分方法（不生成自由文本），保持原始转录文本，在每个词边界做出标点插入决策。该方法在限定的K个子词标记前瞻范围内，将标点插入假设与无插入基线对比，并通过权重α和验证校准阈值τ（推理时不更新参数）进行决策校准。在IWSLT 2017数据集上，无微调设置下（验证校准，K=2）4类宏F1达0.893；微调后（K=2）达0.937，显著优于提示基线（0.566）和微调后的ELECTRA基线（0.913），且保持相同前瞻预算。通过消融实验分析了前瞻预算K的影响。
原文摘要 · Abstract (English)
Punctuation restoration improves ASR (Automatic Speech Recognition) readability. However streaming ASR requires online decisions with limited future context. In streaming ASR, the system predicts punctuation incrementally, which makes generation-based approaches prone to latency and alignment failures under boundary-wise evaluation. This paper proposes a non-autoregressive scoring method (no free-form generation) that preserves the input transcript and makes a decision at each word boundary. Our method compares punctuation insertion hypotheses against a no-insertion baseline under a bounded K-subword-token lookahead, and calibrates decisions using a weight α and a validation-calibrated threshold τ (no parameter updates during inference). On IWSLT 2017, our scoring method achieves a 4-class macro F1 of 0.893 in the no fine-tuning setting (validation-calibrated, K=2) and 0.937 after fine-tuning (K=2), outperforming the prompt-based baseline (0.566) and a fine-tuned ELECTRA baseline (0.913) under the same lookahead budget. We analyze the impact of the lookahead budget through ablation studies on K.
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。