Contrastive Audio-Visual Masked Autoencoder
This work addresses audio-visual event classification and retrieval for multimedia applications, representing an incremental improvement by integrating existing self-supervised frameworks.
The paper tackles the problem of learning joint audio-visual representations by extending Masked Auto-Encoder to multi-modalities and combining it with contrastive learning, achieving a new SOTA accuracy of 65.9% on VGGSound and competitive performance on AudioSet.
In this paper, we first extend the recent Masked Auto-Encoder (MAE) model from a single modality to audio-visual multi-modalities. Subsequently, we propose the Contrastive Audio-Visual Masked Auto-Encoder (CAV-MAE) by combining contrastive learning and masked data modeling, two major self-supervised learning frameworks, to learn a joint and coordinated audio-visual representation. Our experiments show that the contrastive audio-visual correspondence learning objective not only enables the model to perform audio-visual retrieval tasks, but also helps the model learn a better joint representation. As a result, our fully self-supervised pretrained CAV-MAE achieves a new SOTA accuracy of 65.9% on VGGSound, and is comparable with the previous best supervised pretrained model on AudioSet in the audio-visual event classification task. Code and pretrained models are at https://github.com/yuangongnd/cav-mae.