Audio-visual event recognition on AVE (test)
71.64AV AccuracyMaxSim
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MaxSimAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 71.64 | 35.82 | 25.62 | 16.42 | 25.95 | 2.48 | — | — | — | |
| ExFMemAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 71.39 | 44.78 | 28.11 | 10.95 | 27.94 | 4.22 | — | — | — | |
| MinSim+ExFMemAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 71.39 | 58.21 | 29.35 | 26.62 | 38.06 | 14.34 | — | — | — | |
| MinSimAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 70.9 | 57.21 | 25.37 | 21.39 | 34.66 | 10.45 | — | — | — | |
| NoneAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 70.4 | 40.55 | 24.88 | 8.71 | 24.71 | 0 | — | — | — | |
| PMR w/o PERFusion strategy=Concat2022.11 | 67.2 | — | — | — | — | — | — | — | — | |
| PMRFusion strategy=Concat2022.11 | 67.1 | — | — | — | — | — | — | — | — | |
| Modality-dropFusion strategy=Concat2022.11 | 66.4 | — | — | — | — | — | — | — | — | |
| PMRFusion strategy=Film2022.11 | 66.4 | — | — | — | — | — | — | — | — | |
| OGM-GEFusion strategy=Concat2022.11 | 65.9 | — | — | — | — | — | — | — | — | |
| PMR w/o PERFusion strategy=Film2022.11 | 65.6 | — | — | — | — | — | — | — | — | |
| Grad-BlendingFusion strategy=Concat2022.11 | 65.5 | — | — | — | — | — | — | — | — | |
| NaiveFusion strategy=Concat2022.11 | 65.4 | — | — | — | — | — | — | — | — | |
| Grad-BlendingFusion strategy=Film2022.11 | 65.2 | — | — | — | — | — | — | — | — | |
| Unimodal VAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 65.17 | 65.17 | 17.66 | 17.66 | 33.5 | 3.56 | — | — | — | |
| OGM-GEFusion strategy=Film2022.11 | 65.1 | — | — | — | — | — | — | — | — | |
| NaiveFusion strategy=Film2022.11 | 64.4 | — | — | — | — | — | — | — | — | |
| Modality-dropFusion strategy=Film2022.11 | 63.9 | — | — | — | — | — | — | — | — | |
| PCLAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 61.94 | 61.69 | 17.91 | 17.91 | 32.5 | -0.67 | — | — | — | |
| Unimodal AAttack method=FGSM, Attack strength (epsilon)=0.062021.04 | 29.85 | 1.24 | 29.85 | 1.24 | 10.77 | -54.49 | — | — | — | |
| AGMBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 62.94 | 59.55 | |
| ARLBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 65.67 | 62.24 | |
| BaselineBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 61.69 | 58.16 | |
| CPSP2024.04 | — | — | — | — | — | — | 78.6 | — | — | |
| Cross EntropySupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 27.7 | — | |
| Cross EntropySupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 50.08 | — | |
| CSSNet2024.04 | — | — | — | — | — | — | 80.5 | — | — | |
| D&RBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 59.95 | 56.14 | |
| Feature KDSupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 31.01 | — | |
| Feature KDSupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 48.51 | — | |
| Grad-BlendingBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 65.67 | 61.09 | |
| MLABackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 63.18 | 58.92 | |
| MMParetoBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 64.43 | 60.3 | |
| OGM-GEBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 64.68 | 61.33 | |
| PMRBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 64.18 | 59.9 | |
| PSP2024.04 | — | — | — | — | — | — | 77.8 | — | — | |
| TeacherRole=Source Model2026.06 | — | — | — | — | — | — | — | 52.74 | — | |
| TeacherRole=Source Model2026.06 | — | — | — | — | — | — | — | 30.35 | — | |
| TIMModality=V2024.04 | — | — | — | — | — | — | 62.8 | — | — | |
| TIMModality=A2024.04 | — | — | — | — | — | — | 65.5 | — | — | |
| TIMModality=AV2024.04 | — | — | — | — | — | — | 79.2 | — | — | |
| TIMModality=AV, Features=Omni+ASF2024.04 | — | — | — | — | — | — | 79.8 | — | — | |
| UCMKDSupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 34.16 | — | |
| UCMKDSupervision=Unpaired2026.06 | — | — | — | — | — | — | — | 52.24 | — | |
| Vanilla KDSupervision=Paired2026.06 | — | — | — | — | — | — | — | 29.85 | — | |
| Vanilla KDSupervision=Paired2026.06 | — | — | — | — | — | — | — | 49.01 | — | |
| VGMRBackbone=Concatenation2026.06 | — | — | — | — | — | — | — | 65.92 | 61.99 |