Transformers have emerged as a prominent model framework for audio tagging (AT), boasting state-of-the-art (SOTA) performance on the widely-used Audioset dataset. However, their impressive performance often comes at the cost of high memory usage, slow inference speed, and considerable model delay, rendering them impractical for real-world AT applications. In this study, we introduce streaming audio transformers (SAT) that combine the vision transformer (ViT) architecture with Transformer-Xl-like chunk processing, enabling efficient processing of long-range audio signals. Our proposed SAT is benchmarked against other transformer-based SOTA methods, achieving significant improvements in terms of mean average precision (mAP) at a delay of 2s and 1s, while also exhibiting significantly lower memory usage and computational overhead. Checkpoints are publicly available https://github.com/RicherMans/SAT.

翻译：Transformer已成为音频标记任务（AT）中主流的模型框架，在广泛使用的Audioset数据集上展现出最先进的性能（SOTA）。然而，其卓越性能往往以高内存占用、推理速度缓慢以及显著的模型延迟为代价，导致难以应用于实际的AT场景。本研究提出流式音频Transformer（SAT），该方法将视觉Transformer（ViT）架构与类Transformer-Xl的分块处理机制相结合，实现对长范围音频信号的高效处理。我们将所提出的SAT与其他基于Transformer的SOTA方法进行基准测试，在2秒和1秒延迟条件下，平均精度均值（mAP）取得显著提升，同时显示出极低的内存占用和计算开销。模型权重已公开，获取地址为 https://github.com/RicherMans/SAT。