Material Classification on FMD (test)
100Mean AccuracyDual-stream framework
Evaluation results
| Method | Links | |
|---|---|---|
| Dual-stream frameworkClass=fabric2026.03 | 100 | |
| Dual-stream frameworkClass=foliage2026.03 | 97 | |
| GPT-4vClass=wood2026.03 | 95 | |
| Dual-stream frameworkClass=metal2026.03 | 94 | |
| CLIPClass=foliage2026.03 | 93 | |
| Dual-stream frameworkClass=water2026.03 | 92 | |
| CLIPClass=wood2026.03 | 91 | |
| Dual-stream frameworkClass=stone2026.03 | 91 | |
| CLIPClass=plastic2026.03 | 90 | |
| Dual-stream frameworkClass=plastic2026.03 | 90 | |
| Dual-stream frameworkClass=Average2026.03 | 89 | |
| GPT-4vClass=paper2026.03 | 88 | |
| CLIPClass=stone2026.03 | 87 | |
| GPT-4vClass=leather2026.03 | 87 | |
| GPT-4vClass=plastic2026.03 | 87 | |
| Dual-stream frameworkClass=glass2026.03 | 87 | |
| GPT-4vClass=fabric2026.03 | 85 | |
| Dual-stream frameworkClass=leather2026.03 | 85 | |
| CLIPClass=glass2026.03 | 84 | |
| CLIPClass=water2026.03 | 84 | |
| Dual-stream frameworkClass=paper2026.03 | 84 | |
| CLIPClass=paper2026.03 | 83 | |
| GPT-4vClass=water2026.03 | 83 | |
| MatSimClass=foliage2026.03 | 82 | |
| CLIPClass=Average2026.03 | 80 | |
| GPT-4vClass=glass2026.03 | 76 | |
| Dual-stream frameworkClass=wood2026.03 | 75 | |
| GPT-4vClass=Average2026.03 | 74 | |
| CLIPClass=leather2026.03 | 67 | |
| CLIPClass=fabric2026.03 | 66 | |
| GPT-4vClass=stone2026.03 | 64 | |
| CLIPClass=metal2026.03 | 62 | |
| MatSimClass=stone2026.03 | 62 | |
| MatSimClass=wood2026.03 | 62 | |
| GPT-4vClass=metal2026.03 | 61 | |
| MatSimClass=glass2026.03 | 57 | |
| MatSimClass=leather2026.03 | 57 | |
| MatSimClass=water2026.03 | 57 | |
| MatSimClass=Average2026.03 | 56 | |
| MatSimClass=plastic2026.03 | 52 | |
| MatSimClass=fabric2026.03 | 47 | |
| MatSimClass=metal2026.03 | 47 | |
| FastViT-SA12Params=10.60M, FLOPs=1.50G, Pre-training=None2026.02 | 45 | |
| TwistNet-18Params=11.59M, FLOPs=1.85G, Pre-training=None2026.02 | 43.5 | |
| ResNet-18Params=11.20M, FLOPs=1.82G, Pre-training=None2026.02 | 42.6 | |
| SE-ResNet-18Params=11.29M, FLOPs=1.82G, Pre-training=None2026.02 | 40.8 | |
| MatSimClass=paper2026.03 | 39 | |
| RepViT-M1.5Params=13.67M, FLOPs=2.31G, Pre-training=None2026.02 | 36.6 | |
| Swin-TinyParams=27.56M, FLOPs=4.51G, Pre-training=None2026.02 | 35.9 | |
| ConvNeXtV2-NanoParams=15.01M, FLOPs=2.45G, Pre-training=None2026.02 | 29.7 | |
| ConvNeXt-TinyParams=27.86M, FLOPs=4.47G, Pre-training=None2026.02 | 24.3 | |
| GPT-4vClass=foliage2026.03 | 19 |