Active Contrastive Learning of Audio-Visual Video Representations
This work addresses a bottleneck in contrastive learning for audio-visual video representation, offering incremental improvements for tasks like video classification.
The paper tackles the problem of suboptimal representations in contrastive learning due to redundant negative samples by proposing an active contrastive learning approach that builds a diverse and informative dictionary, achieving state-of-the-art performance on audio-visual benchmarks such as UCF101, HMDB51, and ESC50.
Contrastive learning has been shown to produce generalizable representations of audio and visual data by maximizing the lower bound on the mutual information (MI) between different views of an instance. However, obtaining a tight lower bound requires a sample size exponential in MI and thus a large set of negative samples. We can incorporate more samples by building a large queue-based dictionary, but there are theoretical limits to performance improvements even with a large number of negative samples. We hypothesize that \textit{random negative sampling} leads to a highly redundant dictionary that results in suboptimal representations for downstream tasks. In this paper, we propose an active contrastive learning approach that builds an \textit{actively sampled} dictionary with diverse and informative items, which improves the quality of negative samples and improves performances on tasks where there is high mutual information in the data, e.g., video classification. Our model achieves state-of-the-art performance on challenging audio and visual downstream benchmarks including UCF101, HMDB51 and ESC50.\footnote{Code is available at: \url{https://github.com/yunyikristy/CM-ACC}}