Multimodal Classification on UR-FUNNY
64.9AccuracyConFu
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ConFuprotocol=linear-probe2025.11 | 64.9 | — | — | — | |
| GRAMprotocol=linear-probe2025.11 | 64.8 | — | — | — | |
| Symileprotocol=linear-probe2025.11 | 64.7 | — | — | — | |
| MIMOt(s)=0.101± 0.0132025.11 | 64.54 | 52.19 | 50.38 | 62.1 | |
| TRIANGLEprotocol=linear-probe2025.11 | 64.5 | — | — | — | |
| AGMt(s)=0.384± 0.0012025.11 | 64.18 | 54.76 | 49.45 | 62.74 | |
| Tri-CLIPprotocol=linear-probe2025.11 | 64 | — | — | — | |
| MMParetot(s)=0.436± 0.0062025.11 | 63.94 | 52.48 | 50.31 | 63.35 | |
| MGDAt(s)=0.441± 0.0032025.11 | 63.81 | 53.85 | 49.81 | 62.96 | |
| EWt(s)=0.090± 0.0022025.11 | 63.63 | 54 | 49.69 | 62.41 | |
| FACTORCL-SUP(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=true, (X")=false, Fact=true2023.06 | 63.52 | — | — | — | |
| MSLRt(s)=0.074± 0.0122025.11 | 63.16 | 54.77 | 50.69 | 61.89 | |
| MSLt(s)=0.038± 0.0022025.11 | 63.1 | 53.02 | 50.18 | 62.49 | |
| MSESt(s)=0.071± 0.0062025.11 | 62.9 | 53.33 | 49.91 | 62.71 | |
| FACTORCL-SSL(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=true, (X")=true, Fact=true2023.06 | 60.5 | — | — | — | |
| Cross+Self+Fact(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=false, (X")=false, Fact=true2023.06 | 59.91 | — | — | — | |
| OurCL-SSL(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=true, (X")=true, Fact=false2023.06 | 58.79 | — | — | — | |
| OurCL-SUP(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=true, (X")=false, Fact=false2023.06 | 58.32 | — | — | — | |
| Cross+Self(X1; X2)=true, (Xi; X)=true, (X1; X2|Y)=false, (X")=false, Fact=false2023.06 | 56.52 | — | — | — | |
| SupCon(X1; X2)=false, (Xi; X)=true, (X1; X2|Y)=false, (X")=false, Fact=false2023.06 | 50.98 | — | — | — | |
| SimCLR(X1; X2)=false, (Xi; X)=true, (X1; X2|Y)=false, (X")=false, Fact=false2023.06 | 50.09 | — | — | — | |
| Audiot(s)=0.037± 0.0022025.11 | — | 58 | — | — | |
| Textt(s)=0.038± 0.0022025.11 | — | — | — | 63.74 | |
| Visualt(s)=0.038± 0.0022025.11 | — | — | 53.12 | — |