Efficient Training of Audio Transformers with Patchout

Citations Over TimeTop 1% of 2022 papers

Abstract

The great success of transformer-based models in natural language processing (NLP) has led to various attempts at adapting these architectures to other domains such as vision and audio. Recent work has shown that transformers can outperform Convolutional Neural Networks (CNNs) on vision and audio tasks. However, one of the main shortcomings of transformer models, compared to the well-established CNNs, is the computational complexity. In transformers, the compute and memory complexity is known to grow quadratically with the input length. Therefore, there has been extensive work on optimizing transformers, but often at the cost of degrading predictive performance. In this work, we propose a novel method to optimize and regularize transformers on audio spectrograms. Our proposed models achieve a new state-of-the-art performance on Audioset and can be trained on a single consumer-grade GPU. Furthermore, we propose a transformer model that outperforms CNNs in terms of both performance and training speed. Source code: https://github.com/kkoutini/PaSST

Related Papers

• A measure for audio-visual programs in schools : prepared for the Committee on Visual Aids in Education(1944)
• A comparative cost-effectiveness study of quotlow-cost audio-visual teaching aids and quothigh-costquot audio-visual teaching aids(1984)
• CAI Yuan-pei:The Promoter of China Early Audio-visual Education(2011)
• Extent of utilization and user’s satisfaction of the library and audio visual center services and programs of the University of St. La Salle - Integrated School libraries and audio - visual centers(2010)