Multiscale Multimodal Transformer for Multimodal Action Recognition

Benchmark Model Rank Results
action-recognition-on-epic-kitchens-100MMT–Action@1: 47.8Verb@1: 70.1Noun@1: 61.0
audio-classification-on-vggsoundMMT (Audio-Visual)–Top 1 Accuracy: 66.2Top 5 Accuracy: 85.7
audio-classification-on-vggsoundMMT (Video)–Top 1 Accuracy: 56.1Top 5 Accuracy: 77.9
multi-modal-classification-on-vgg-soundMMT–Top-1 Accuracy: 66.2Top-5 Accuracy: 85.7