RALAD: Bridging the Real-to-Sim Domain Gap in Autonomous Driving with Retrieval-Augmented Learning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913954484191232 |
|---|---|
| author | Zuo, Jiacheng Hu, Haibo Zhou, Zikang Cui, Yufei Liu, Ziquan Wang, Jianping Guan, Nan Wang, Jin Xue, Chun Jason |
| author_facet | Zuo, Jiacheng Hu, Haibo Zhou, Zikang Cui, Yufei Liu, Ziquan Wang, Jianping Guan, Nan Wang, Jin Xue, Chun Jason |
| contents | In the pursuit of robust autonomous driving systems, models trained on real-world datasets often struggle to adapt to new environments, particularly when confronted with corner cases such as extreme weather conditions. Collecting these corner cases in the real world is non-trivial, which necessitates the use of simulators for validation. However,the high computational cost and the domain gap in data distribution have hindered the seamless transition between real and simulated driving scenarios. To tackle this challenge, we propose Retrieval-Augmented Learning for Autonomous Driving (RALAD), a novel framework designed to bridge the real-to-sim gap at a low cost. RALAD features three primary designs, including (1) domain adaptation via an enhanced Optimal Transport (OT) method that accounts for both individual and grouped image distances, (2) a simple and unified framework that can be applied to various models, and (3) efficient fine-tuning techniques that freeze the computationally expensive layers while maintaining robustness. Experimental results demonstrate that RALAD compensates for the performance degradation in simulated environments while maintaining accuracy in real-world scenarios across three different models. Taking Cross View as an example, the mIOU and mAP metrics in real-world scenarios remain stable before and after RALAD fine-tuning, while in simulated environments,the mIOU and mAP metrics are improved by 10.30% and 12.29%, respectively. Moreover, the re-training cost of our approach is reduced by approximately 88.1%. Our code is available at https://github.com/JiachengZuo/RALAD.git. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_12296 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RALAD: Bridging the Real-to-Sim Domain Gap in Autonomous Driving with Retrieval-Augmented Learning Zuo, Jiacheng Hu, Haibo Zhou, Zikang Cui, Yufei Liu, Ziquan Wang, Jianping Guan, Nan Wang, Jin Xue, Chun Jason Computer Vision and Pattern Recognition Artificial Intelligence In the pursuit of robust autonomous driving systems, models trained on real-world datasets often struggle to adapt to new environments, particularly when confronted with corner cases such as extreme weather conditions. Collecting these corner cases in the real world is non-trivial, which necessitates the use of simulators for validation. However,the high computational cost and the domain gap in data distribution have hindered the seamless transition between real and simulated driving scenarios. To tackle this challenge, we propose Retrieval-Augmented Learning for Autonomous Driving (RALAD), a novel framework designed to bridge the real-to-sim gap at a low cost. RALAD features three primary designs, including (1) domain adaptation via an enhanced Optimal Transport (OT) method that accounts for both individual and grouped image distances, (2) a simple and unified framework that can be applied to various models, and (3) efficient fine-tuning techniques that freeze the computationally expensive layers while maintaining robustness. Experimental results demonstrate that RALAD compensates for the performance degradation in simulated environments while maintaining accuracy in real-world scenarios across three different models. Taking Cross View as an example, the mIOU and mAP metrics in real-world scenarios remain stable before and after RALAD fine-tuning, while in simulated environments,the mIOU and mAP metrics are improved by 10.30% and 12.29%, respectively. Moreover, the re-training cost of our approach is reduced by approximately 88.1%. Our code is available at https://github.com/JiachengZuo/RALAD.git. |
| title | RALAD: Bridging the Real-to-Sim Domain Gap in Autonomous Driving with Retrieval-Augmented Learning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2501.12296 |