CMAE-V: Contrastive Masked Autoencoders for Video Action Recognition
This is an incremental improvement for video action recognition, applying an existing self-supervised method to a new domain with strong performance gains.
The paper tackled video action recognition by adapting the Contrastive Masked Autoencoder (CMAE) framework to videos without architectural changes, achieving 82.2% top-1 accuracy on Kinetics-400 and 71.6% on Something-something V2.
Contrastive Masked Autoencoder (CMAE), as a new self-supervised framework, has shown its potential of learning expressive feature representations in visual image recognition. This work shows that CMAE also trivially generalizes well on video action recognition without modifying the architecture and the loss criterion. By directly replacing the original pixel shift with the temporal shift, our CMAE for visual action recognition, CMAE-V for short, can generate stronger feature representations than its counterpart based on pure masked autoencoders. Notably, CMAE-V, with a hybrid architecture, can achieve 82.2% and 71.6% top-1 accuracy on the Kinetics-400 and Something-something V2 datasets, respectively. We hope this report could provide some informative inspiration for future works.