Fine-grained Image Classification on Combined Biodiversity Benchmarks Weighted Avg
70.8Weighted Avg AccuracyBT-Clip-O
Evaluation results
| Method | Links | |
|---|---|---|
| BT-Clip-OBackbone=ViT-B-16, Pre-training=Fine-tuned from OpenAI-B, Evaluation Protocol=Fine-tuned2024.06 | 70.8 | |
| BT-Clip-MBackbone=ViT-L-14, Pre-training=Fine-tuned from MetaCLIP-L, Evaluation Protocol=Fine-tuned2024.06 | 69.5 | |
| BT-Clip-BBackbone=ViT-B-16, Pre-training=Fine-tuned from BioCLIP-B, Evaluation Protocol=Fine-tuned2024.06 | 67.2 | |
| BioCLIP-BBackbone=ViT-B-16, Pre-training=BioCLIP, Evaluation Protocol=Zero-shot2024.06 | 58.5 | |
| MetaCLIP-LBackbone=ViT-L-14, Pre-training=MetaCLIP-cc, Evaluation Protocol=Zero-shot2024.06 | 25 | |
| OpenAI-BBackbone=ViT-B-16, Pre-training=OpenAI, Evaluation Protocol=Zero-shot2024.06 | 14.7 |