Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.24922 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912798160715776 |
|---|---|
| author | Olber, Bartłomiej Winter, Jakub Wawrzyński, Paweł Gamalii, Andrii Górniak, Daniel Łojek, Marcin Nowak, Robert Radlak, Krystian |
| author_facet | Olber, Bartłomiej Winter, Jakub Wawrzyński, Paweł Gamalii, Andrii Górniak, Daniel Łojek, Marcin Nowak, Robert Radlak, Krystian |
| contents | 3D object detectors are fundamental components of perception systems in autonomous vehicles. While these detectors achieve remarkable performance on standard autonomous driving benchmarks, they often struggle to generalize across different domains - for instance, a model trained in the U.S. may perform poorly in regions like Asia or Europe. This paper presents a novel lidar domain adaptation method based on neuron activation patterns, demonstrating that state-of-the-art performance can be achieved by annotating only a small, representative, and diverse subset of samples from the target domain if they are correctly selected. The proposed approach requires very small annotation budget and, when combined with post-training techniques inspired by continual learning prevent weight drift from the original model. Empirical evaluation shows that the proposed domain adaptation approach outperforms both linear probing and state-of-the-art domain adaptation techniques. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_24922 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Semi-Supervised Diversity-Aware Domain Adaptation for 3D Object detection Olber, Bartłomiej Winter, Jakub Wawrzyński, Paweł Gamalii, Andrii Górniak, Daniel Łojek, Marcin Nowak, Robert Radlak, Krystian Computer Vision and Pattern Recognition 3D object detectors are fundamental components of perception systems in autonomous vehicles. While these detectors achieve remarkable performance on standard autonomous driving benchmarks, they often struggle to generalize across different domains - for instance, a model trained in the U.S. may perform poorly in regions like Asia or Europe. This paper presents a novel lidar domain adaptation method based on neuron activation patterns, demonstrating that state-of-the-art performance can be achieved by annotating only a small, representative, and diverse subset of samples from the target domain if they are correctly selected. The proposed approach requires very small annotation budget and, when combined with post-training techniques inspired by continual learning prevent weight drift from the original model. Empirical evaluation shows that the proposed domain adaptation approach outperforms both linear probing and state-of-the-art domain adaptation techniques. |
| title | Semi-Supervised Diversity-Aware Domain Adaptation for 3D Object detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.24922 |