Gender Classification on DiveFace (test)
97.371Accuracy (East Asian Male)CLIP
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| CLIPText Generation Source=BLIP Image-to-Text Model (T1)2025.12 | 97.371 | 95.136 | 98.359 | 82.793 | 99.426 | 90.136 | 1.2 | 93.867 | 5.795 | |
| ITM GuidedText Generation Source=Attributes (T2)2025.12 | 97.189 | 97.195 | 97.31 | 89.865 | 99.58 | 93 | 1.108 | 95.694 | 3.248 | |
| CLIPText Generation Source=Attributes (T2)2025.12 | 97 | 92.682 | 98.5 | 76.892 | 99.337 | 86.438 | 1.292 | 91.798 | 7.964 | |
| TandemNetText Generation Source=Attributes (T2)2025.12 | 96.6 | 95.924 | 97.174 | 88.243 | 99.426 | 92.162 | 1.127 | 94.924 | 3.683 | |
| Image Only2025.12 | 96.374 | 96.231 | 97.675 | 86.67 | 99.116 | 92.426 | 1.143 | 94.753 | 4.147 | |
| Image-Text FusionText Generation Source=BLIP Image-to-Text Model (T1)2025.12 | 95.83 | 98.948 | 96 | 96.306 | 99.028 | 97.094 | 1.033 | 97.218 | 1.326 | |
| TandemNetText Generation Source=BLIP Image-to-Text Model (T1)2025.12 | 95.648 | 99.167 | 95.488 | 96.532 | 98.895 | 97.05 | 1.039 | 97.149 | 1.444 | |
| ITM GuidedText Generation Source=BLIP Image-to-Text Model (T1)2025.12 | 95.33 | 99.036 | 95.168 | 96.62 | 99.116 | 96.61 | 1.042 | 97 | 1.584 | |
| Image-Text FusionText Generation Source=Attributes (T2)2025.12 | 95.014 | 98.773 | 96.08 | 94.68 | 99.116 | 95.773 | 1.047 | 96.591 | 1.741 |