arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2401.08415cs.SDcs.LGeess.AS
收藏

从粗到细:音频频谱图Transformer的高效训练

From Coarse to Fine: Efficient Training for Audio Spectrogram Transformers

Korea Advanced Institute of Science and Technology(韩国科学技术院)

Jiu Feng, Mehmet Hamza Erol, Joon Son Chung, Arda Senocak

AI总结:

本文提出基于时间压缩的粗到细多阶段课程训练方法,在降低计算资源和时间成本的同时提升或保持AST音频分类性能,并适用于其他AST方法。

摘要:

Transformer已成为近期音频分类进展的核心。然而,从头训练音频频谱图Transformer(例如AST)可能耗费大量资源和时间。此外,Transformer的复杂度在很大程度上取决于输入音频频谱图的尺寸。在本工作中,我们旨在通过关联时间轴上的分辨率来优化AST训练。我们通过将粗到细的经典思想与Transformer模型相结合,引入音频频谱图Transformer的多阶段训练。为实现这一目标,我们提出了一组时间压缩方法。通过采用其中一种方法,Transformer模型在初始阶段从低分辨率(粗粒度)数据中学习,随后在课程学习策略下使用高分辨率数据进行微调。实验结果表明,所提出的AST训练机制能够实现性能提升(或持平)并加快收敛,即需要更少的计算资源和更少的时间。该方法还可推广到其他基于AST的方法,无论其学习范式如何。

英文摘要:

Transformers have become central to recent advances in audio classification. However, training an audio spectrogram transformer, e.g. AST, from scratch can be resource and time-intensive. Furthermore, the complexity of transformers heavily depends on the input audio spectrogram size. In this work, we aim to optimize AST training by linking to the resolution in the time-axis. We introduce multi-phase training of audio spectrogram transformers by connecting the seminal idea of coarse-to-fine with transformer models. To achieve this, we propose a set of methods for temporal compression. By employing one of these methods, the transformer model learns from lower-resolution (coarse) data in the initial phases, and then is fine-tuned with high-resolution data later in a curriculum learning strategy. Experimental results demonstrate that the proposed training mechanism for AST leads to improved (or on-par) performance with faster convergence, i.e. requiring fewer computational resources and less time. This approach is also generalizable to other AST-based methods regardless of their learning paradigms.

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。

↑