We present Masked Audio-Video Learners (MAViL) to train audio-visual representations. Our approach learns with three complementary forms of self-supervision: (1) reconstruction of masked audio and video input data, (2) intra- and inter-modal contrastive learning with masking, and (3) self-training by reconstructing joint audio-video contextualized features learned from the first two objectives. Pre-training with MAViL not only enables the model to perform well in audio-visual classification and retrieval tasks but also improves representations of each modality in isolation, without using information from the other modality for fine-tuning or inference. Empirically, MAViL sets a new state-of-the-art on AudioSet (53.1 mAP) and VGGSound (67.1% accuracy). For the first time, a self-supervised audio-visual model outperforms ones that use external supervision on these benchmarks.
@article{arxiv.2212.08071,
title = {MAViL: Masked Audio-Video Learners},
author = {Po-Yao Huang and Vasu Sharma and Hu Xu and Chaitanya Ryali and Haoqi Fan and Yanghao Li and Shang-Wen Li and Gargi Ghosh and Jitendra Malik and Christoph Feichtenhofer},
journal= {arXiv preprint arXiv:2212.08071},
year = {2023}
}