Leveraging Confident Image Regions for Source-Free Domain-Adaptive Object Detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911191420370944 |
|---|---|
| author | Mekhalfi, Mohamed Lamine Boscaini, Davide Poiesi, Fabio |
| author_facet | Mekhalfi, Mohamed Lamine Boscaini, Davide Poiesi, Fabio |
| contents | Source-free domain-adaptive object detection is an interesting but scarcely addressed topic. It aims at adapting a source-pretrained detector to a distinct target domain without resorting to source data during adaptation. So far, there is no data augmentation scheme tailored to source-free domain-adaptive object detection. To this end, this paper presents a novel data augmentation approach that cuts out target image regions where the detector is confident, augments them along with their respective pseudo-labels, and joins them into a challenging target image to adapt the detector. As the source data is out of reach during adaptation, we implement our approach within a teacher-student learning paradigm to ensure that the model does not collapse during the adaptation procedure. We evaluated our approach on three adaptation benchmarks of traffic scenes, scoring new state-of-the-art on two of them. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_10081 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Leveraging Confident Image Regions for Source-Free Domain-Adaptive Object Detection Mekhalfi, Mohamed Lamine Boscaini, Davide Poiesi, Fabio Computer Vision and Pattern Recognition Source-free domain-adaptive object detection is an interesting but scarcely addressed topic. It aims at adapting a source-pretrained detector to a distinct target domain without resorting to source data during adaptation. So far, there is no data augmentation scheme tailored to source-free domain-adaptive object detection. To this end, this paper presents a novel data augmentation approach that cuts out target image regions where the detector is confident, augments them along with their respective pseudo-labels, and joins them into a challenging target image to adapt the detector. As the source data is out of reach during adaptation, we implement our approach within a teacher-student learning paradigm to ensure that the model does not collapse during the adaptation procedure. We evaluated our approach on three adaptation benchmarks of traffic scenes, scoring new state-of-the-art on two of them. |
| title | Leveraging Confident Image Regions for Source-Free Domain-Adaptive Object Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2501.10081 |