Multimodal Retrieval on Ele-fashion MAG benchmark
76.3R@1GOMA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GOMAMethod Category=Our Method, Feature Backbone=Qwen2-VL2026.05 | 76.3 | 94.6 | 97.3 | |
| DMGCMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 43.2 | 57.8 | 63.5 | |
| DGFMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 30.8 | 53.6 | 62.6 | |
| SmartCLIPMethod Category=Pairwise Image-Text Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 25.7 | 55.9 | 69 | |
| CLIP-RefineMethod Category=Pairwise Image-Text Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 22.5 | 51.3 | 64.4 | |
| D2S-VSEMethod Category=Pairwise Image-Text Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 15.6 | 37.6 | 49.3 | |
| LGMRecMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 15.3 | 39.2 | 52.6 | |
| LinearMethod Category=Direct Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 14.2 | 34.9 | 46.8 | |
| VSE++Method Category=Pairwise Image-Text Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 13.4 | 32.7 | 43.2 | |
| MMGCNMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 9.5 | 28 | 40.5 | |
| MGATMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 5.5 | 16.7 | 25 | |
| GSMNMethod Category=Multimodal Graph Retrieval & Recent MAG Models, Feature Backbone=Qwen2-VL2026.05 | 3.2 | 12.4 | 20.3 | |
| RawMethod Category=Direct Retrieval Baselines, Feature Backbone=Qwen2-VL2026.05 | 0.9 | 2.5 | 3.8 |