Learning Correlation Structures for Vision Transformers

Benchmark Model Rank Results
action-classification-on-kinetics-400StructViT-B-4-1Acc@1: 83.4
action-recognition-in-videos-on-somethingStructVit-B-4-1Top-1 Accuracy: 71.5
action-recognition-in-videos-on-something-1StructVit-B-4-1Top 1 Accuracy: 61.3
action-recognition-on-diving-48StructVit-B-4-1Accuracy: 88.3