Text-to-Depth Retrieval (via Vision pivot) on NYU to SUN (test)
33.1mAPIB+BB
Evaluation results
| Method | Links | |
|---|---|---|
| IB+BBBackbone=ImageBind, Variant=BrokenBind2026.02 | 33.1 | |
| LB+BBBackbone=LanguageBind, Variant=BrokenBind2026.02 | 29.4 | |
| LB+FTBackbone=LanguageBind, Variant=Fine-tuned2026.02 | 13.2 | |
| IB+FTBackbone=ImageBind, Variant=Fine-tuned2026.02 | 12.5 | |
| LBBackbone=LanguageBind, Variant=Zero-shot2026.02 | 4.1 | |
| IBBackbone=ImageBind, Variant=Zero-shot2026.02 | 3.9 |