Masked Feature Modelling for the unsupervised pre-training of a Graph Attention Network block for bottom-up video event recognition.

Dimitrios Daskalakis, Nikolaos Gkalelis, Vasileios Mezaris

Browse the full ISM paper archive.