MCCLA: An improved contrastive learning structure for audio representations
Published in arXiv, 2023
This paper proposes MCCLA, an improved contrastive learning structure designed for learning robust audio representations. By enhancing the contrastive learning framework with multi-scale features, our approach achieves better audio representation quality for downstream tasks.
Recommended citation: @article{wang2023mccla, title={MCCLA: An improved contrastive learning structure for audio representations}, author={Wang, Yanwen and Shah, Ankit and Qiu, Xunwen and Cao, Yanran and Xu, Yizi and Raj, Bhiksha}, journal={arXiv}, year={2023} }