Mobile Agent Evaluation on AndroidControl High (test)
43.16GroundingAria-UITH
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Aria-UITHEvaluation Protocol=W. Training Set, Planner=GPT-4o2024.12 | 43.16 | 10.17 | |
| Aria-UIIHEvaluation Protocol=W. Training Set, Planner=GPT-4o2024.12 | 42.97 | 10.1 | |
| Aria-UIEvaluation Protocol=W. Training Set, Planner=GPT-4o2024.12 | 41.78 | 9.97 | |
| UGroundEvaluation Protocol=W. Training Set, Planner=GPT-4o2024.12 | 37.98 | 9.15 | |
| Aria-UIEvaluation Protocol=Zero-shot, Planner=GPT-4o2024.12 | 35.12 | 5.95 | |
| Qwen2-VLEvaluation Protocol=Zero-shot, Planner=GPT-4o2024.12 | 30.32 | 4.08 | |
| SeeClickEvaluation Protocol=Zero-shot, Planner=GPT-4o2024.12 | 20.17 | 4.29 | |
| GPT-4oEvaluation Protocol=Zero-shot, Planner=GPT-4o2024.12 | 10.36 | 2.84 |