CNext-trans

Reported on 7 benchmarks across 1 task

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

Audio7 results

Audio captioningonAudioCaps
#params (M)
40
Audio captioningonAudioCaps
CIDEr
0.8061
best: 50.2 (Audio Flamingo)
Audio captioningonAudioCaps
FENSE
0.6431
best: 0.668 (SLAM-AAC)
Audio captioningonAudioCaps
METEOR
0.2527
best: 20.5 (Audio Flamingo)
Audio captioningonAudioCaps
SPICE
0.1841
best: 15.1 (Audio Flamingo)
Audio captioningonAudioCaps
SPIDEr
0.4951
best: 32.6 (Audio Flamingo)
Audio captioningonAudioCaps
SPIDEr-FL
0.4945
best: 0.515 (SLAM-AAC)