通过捕捉语音识别中帧的尖峰特征，大幅加速解码同时保持精度。
Breaking Through the Spike: Spike Window Decoding for Accelerated and Precise Automatic Speech Recognition
- 基于CTC输出的尖峰特性，只解码关键帧，减少计算量。
- 在AISHELL-1和内部大数据集上实现最优准确率，推理速度显著提升。
- 适合需要高速高精度语音识别的工业级应用，如实时转写系统。
近年来，端到端自动语音识别已成为产业与学术界的主流方法。为优化特定场景下的系统性能，加权有限状态转换器（WFST）被广泛用于融合声学模型与语言模型，其静态图结构可隐式集成语言模型，保障识别鲁棒性并支持快速纠错。然而，WFST需逐帧对CTC后验概率进行自回归搜索，严重制约推理速度。本文深入研究了CTC输出的尖峰特性，提出假设：非空帧相邻的尖峰帧携带语义信息，对模型有益。基于此，提出尖峰窗口解码（Spike Window Decoding）算法，使WFST解码的帧数与CTC输出中的尖峰帧数呈线性关系，极大提升推理速度，同时保证识别性能。该方法在AISHELL-1及大规模内部数据集上均取得当前最优准确率，显著加速解码，开创了将CTC输出与WFST融合的新范式。
原文摘要 · Abstract (English)
Recently, end-to-end automatic speech recognition has become the mainstream approach in both industry and academia. To optimize system performance in specific scenarios, the Weighted Finite-State Transducer (WFST) is extensively used to integrate acoustic and language models, leveraging its capacity to implicitly fuse language models within static graphs, thereby ensuring robust recognition while also facilitating rapid error correction. However, WFST necessitates a frame-by-frame search of CTC posterior probabilities through autoregression, which significantly hampers inference speed. In this work, we thoroughly investigate the spike property of CTC outputs and further propose the conjecture that adjacent frames to non-blank spikes carry semantic information beneficial to the model. Building on this, we propose the Spike Window Decoding algorithm, which greatly improves the inference speed by making the number of frames decoded in WFST linearly related to the number of spiking frames in the CTC output, while guaranteeing the recognition performance. Our method achieves SOTA recognition accuracy with significantly accelerates decoding speed, proven across both AISHELL-1 and large-scale In-House datasets, establishing a pioneering approach for integrating CTC output with WFST.
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。