UI Screen Classification on RICO SCA
93.3AccuracySigLIP 2
Evaluation Results
| Method | Links | |
|---|---|---|
| SigLIP 2LLM Decoder=Llama-3.2-3B-Instruct2025.12 | 93.3 | |
| DAVELLM Decoder=Llama-3.2-3B-Instruct2025.12 | 92.8 | |
| AIMv2LLM Decoder=Llama-3.2-3B-Instruct2025.12 | 91.3 | |
| Pix2StructLLM Decoder=Llama-3.2-3B-Instruct2025.12 | 90.1 | |
| DolphinLLM Decoder=Llama-3.2-3B-Instruct2025.12 | 88.4 | |
| MAE-WebLLM Decoder=Llama-3.2-3B-Instruct2025.12 | 88.3 | |
| DinoV2LLM Decoder=Llama-3.2-3B-Instruct2025.12 | 85.6 |