Contextualized Video Instance Retrieval on This-Is-My test-time personalization dataset P
4,200MRRMeta-Personalized Vision-Language Model
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Meta-Personalized Vision-Language Model2023.06 | 4,200 | 5,070 | |
| CLIP (language)Modality=language2023.06 | 3,080 | 3,670 | |
| CLIP (V+L)Modality=visual + language2023.06 | 2,090 | 2,330 | |
| CLIP (visual)Modality=visual2023.06 | 1,030 | 1,200 | |
| Random2023.06 | 0 | 0 |