Zero Shot Composed Image Retrieval Zs Cir On 2
評価指標
(Recall@10+Recall@50)/2
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | (Recall@10+Recall@50)/2 |
|---|---|
| reducing-task-discrepancy-of-text-encoders | 56.74 |
| training-free-zs-cir-via-weighted-modality | 47.16 |
| zero-shot-composed-image-retrieval-with | 32.71 |
| zero-shot-composed-image-retrieval-with | 37.76 |
| language-only-efficient-training-of-zero-shot | 55.40 |
| isearle-improving-textual-inversion-for-zero | 38.24 |
| compodiff-versatile-composed-image-retrieval | 45.37 |
| zero-shot-composed-image-retrieval-with | 35.90 |
| context-i2w-mapping-images-to-context | 38.35 |
| training-free-zs-cir-via-weighted-modality | 44.58 |
| magiclens-self-supervised-image-retrieval | 45.3 |
| training-free-zs-cir-via-weighted-modality | 41.27 |
| training-free-zs-cir-via-weighted-modality | 39.84 |
| magiclens-self-supervised-image-retrieval | 48.1 |
| vision-by-language-for-training-free | 38.56 |
| collm-a-large-language-model-for-composed | 39.8 |
| reason-before-retrieve-one-stage-reflective | 42.87 |
| this-is-my-unicorn-fluffy-personalizing | 28.51 |
| covr-learning-composed-video-retrieval-from | 48.3 |
| reason-before-retrieve-one-stage-reflective | 47.34 |
| zero-shot-composed-image-retrieval-with | 32.39 |
| reducing-task-discrepancy-of-text-encoders | 40.66 |
| pretrain-like-you-inference-masked-tuning | 46.42 |
| imagescope-unifying-language-guided-image-1 | - |
| reason-before-retrieve-one-stage-reflective | 42.82 |
| zero-shot-composed-text-image-retrieval | 44.75 |
| pic2word-mapping-pictures-to-words-for-zero | 34.20 |
| magiclens-self-supervised-image-retrieval | 41.6 |
| semantic-editing-increment-benefits-zero-shot | 54.45 |
| collm-a-large-language-model-for-composed | 45.3 |
| isearle-improving-textual-inversion-for-zero | 34.93 |
| language-only-efficient-training-of-zero-shot | 36.39 |
| ldre-llm-based-divergent-reasoning-and | 43.98 |
| magiclens-self-supervised-image-retrieval | 36.85 |
| isearle-improving-textual-inversion-for-zero | 39.39 |
| collm-a-large-language-model-for-composed | 49.9 |
| isearle-improving-textual-inversion-for-zero | 34.60 |
| vision-by-language-for-training-free | 42.28 |
| vision-by-language-for-training-free | 38.82 |
| compodiff-versatile-composed-image-retrieval | 44.11 |