DriveLM
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Autonomous Driving Reasoning | DriveLM-ADvLM | Final Score49.89 | 24 | |
| Driving with language reasoning | DriveLM (leaderboard) | Accuracy83 | 12 | |
| Autonomous driving reasoning (cross-view risk object perception, action prediction, and planning) | DriveLM | Reasoning Score (P/P/P)61.3 | 12 | |
| Driving VQA | DriveLM (test) | Accuracy81 | 11 | |
| Autonomous Driving Evaluation | DriveLM VRU-Accident benchmark | GAR64.13 | 10 | |
| Visual Question Answering | DriveLM | BLEU-454.56 | 8 | |
| Graph Visual Question Answering | DriveLM GVQA | Accuracy74 | 7 | |
| Vision-Language Autonomous Driving | DriveLM open-loop | Accuracy73.81 | 6 | |
| Autonomous Driving | DriveLM | Description Score30 | 6 | |
| Graph VQA | DriveLM (test) | BLEU-459.01 | 6 | |
| Language Understanding | DriveLM | BLEU-453.09 | 6 | |
| Driving Question Answering | DriveLM | Accuracy81.23 | 5 | |
| Autonomous Driving Reasoning | DriveLM 25% Token Retention | Accuracy81.23 | 5 | |
| Reasoning and Generation | DriveLM (test) | Accuracy81 | 5 | |
| Driving Scene Understanding | DriveLM (test) | SPICE45.45 | 5 | |
| Generative Question Answering | DriveLM (test) | BLEU-453.09 | 5 | |
| Drive VQA | DriveLM | GPT Score67.3 | 5 | |
| Drive Scene Understanding | DriveLM | Params (B)0.9 | 3 | |
| Spatial Perception | DriveLM | Match Score13.43 | 3 |
Showing 19 of 19 rows