Image Retrieval on xFlickr&CO
62.98Recall@1ToMCLIP(Lta)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ToMCLIP(Lta)Setting=Full data (2M samples), Image Encoder=ViT-B/16+2025.10 | 62.98 | 85.79 | 91.6 | |
| MCLIPSetting=Full data (2M samples), Image Encoder=ViT-B/16+2025.10 | 62.24 | 85.27 | 91.09 | |
| ToMCLIP(Ldm)Setting=Full data (2M samples), Image Encoder=ViT-B/16+2025.10 | 62.24 | 85.39 | 91.22 | |
| ToMCLIPSetting=Full data (2M samples), Image Encoder=ViT-B/16+2025.10 | 61.91 | 84.89 | 90.78 | |
| ToMCLIPSetting=Low resource (1% subset), Image Encoder=ViT-B/16+2025.10 | 58.53 | 83.37 | 90.51 | |
| ToMCLIP(Ldm)Setting=Low resource (1% subset), Image Encoder=ViT-B/16+2025.10 | 57.91 | 83.15 | 90.37 | |
| ToMCLIP(Lta)Setting=Low resource (1% subset), Image Encoder=ViT-B/16+2025.10 | 57.58 | 82.77 | 90.12 | |
| MCLIPSetting=Low resource (1% subset), Image Encoder=ViT-B/16+2025.10 | 56.44 | 82.28 | 89.6 | |
| CLIPSetting=Low resource (1% subset), Image Encoder=ViT-B/16+2025.10 | 16.38 | 27 | 32.06 | |
| CLIPSetting=Full data (2M samples), Image Encoder=ViT-B/16+2025.10 | 16.38 | 27 | 32.06 |