MyVLM
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Personalized Grounding | MyVLM | Precision100 | 14 | |
| Visual recognition | MyVLM | Positive Accuracy98.1 | 11 | |
| Recognition | MyVLM | Single Recall98.4 | 11 | |
| Image Captioning | MyVLM | Caption Recall (Single)0.975 | 11 | |
| Captioning | MyVLM Dataset | Precision0.957 | 5 | |
| Retrieval | MyVLM | Hit Rate@192.2 | 5 | |
| Captioning | MyVLM Single Concept (test) | Recall91.3 | 4 | |
| Recognition | MyVLM Single Concept, 1 Reference View | Precision86 | 4 | |
| instance-aware caption generation | MyVLM benchmark | CLIP Image Similarity27.06 | 4 | |
| Recognition | MyVLM Single Concept, 5 Reference Views 1 | Precision87.7 | 3 | |
| Instance Identification | MyVLM benchmark | Positive Score97 | 3 | |
| Captioning | MyVLM | CLIPScore30.2 | 2 |
Showing 12 of 12 rows