Semi-Supervised Hard Attention (SSHA); pretrained on Deepmind Kinetics dataset

Reported on 1 benchmark across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.