Skip to main content
. 2024 Apr 1;19(4):e0301098. doi: 10.1371/journal.pone.0301098

Fig 5. AVMIT vs other datasets.

Fig 5

Audiovisual similarity scores, as estimated by MMV, across a series of audiovisual action recognition datasets; AVMIT (ours), MIT-16, Kinetics-Sounds, VGG-Sound and AVE. (a) Average audiovisual similarity score across entire datasets. (b) Rain cloud plot showing the distribution of audiovisual similarity scores for each dataset.