UAVM: Towards Unifying Audio and Visual Models

Benchmark Model Rank Results
multi-modal-classification-on-vgg-soundUAVM#2Top-1 Accuracy: 65.8
audio-classification-on-audiosetUAVM (Audio + Video)#5Test mAP: 0.504
audio-classification-on-vggsoundUAVM (Audio + Video)#5Top 1 Accuracy: 65.8
audio-classification-on-vggsoundUAVM (Audio Only)#11Top 1 Accuracy: 56.5
audio-classification-on-vggsoundUAVM (Video Only)#16Top 1 Accuracy: 49.9