{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/the-role-of-pre-training-in-high-resolution","title":"Do we still need ImageNet pre-training in remote sensing scene classification?","arxiv_id":"2111.03690","date":"2021-11-05","proceeding":null,"authors":["Vladimir Risojević","Vladan Stojnić"],"abstract":"Due to the scarcity of labeled data, using supervised models pre-trained on ImageNet is a de facto standard in remote sensing scene classification. Recently, the availability of larger high resolution remote sensing (HRRS) image datasets and progress in self-supervised learning have brought up the questions of whether supervised ImageNet pre-training is still necessary for remote sensing scene classification and would supervised pre-training on HRRS image datasets or self-supervised pre-training on ImageNet achieve better results on target remote sensing scene classification tasks. To answer these questions, in this paper we both train models from scratch and fine-tune supervised and self-supervised ImageNet models on several HRRS image datasets. We also evaluate the transferability of learned representations to HRRS scene classification tasks and show that self-supervised pre-training outperforms the supervised one, while the performance of HRRS pre-training is similar to self-supervised pre-training or slightly lower. Finally, we propose using an ImageNet pre-trained model combined with a second round of pre-training using in-domain HRRS images, i.e. domain-adaptive pre-training. The experimental results show that domain-adaptive pre-training results in models that achieve state-of-the-art results on HRRS scene classification benchmarks. The source code and pre-trained models are available at \\url{https://github.com/risojevicv/RSSC-transfer}.","url_abs":"https://arxiv.org/abs/2111.03690v3","url_pdf":"https://arxiv.org/pdf/2111.03690v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"the-role-of-pre-training-in-high-resolution","repo_url":"https://github.com/risojevicv/rssc-transfer","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"multi-label-classification","task_name":"Multi-Label Classification"},{"task_slug":"scene-classification","task_name":"Scene Classification"},{"task_slug":"self-supervised-learning","task_name":"Self-Supervised Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multi-label-classification-on-mlrsnet","task":"Multi-Label Classification","dataset":"MLRSNet","model":"ResNet50 (fine-tuning)","rank_in_archive_order":1,"of":2,"metrics":{"F1-score":"92.41"},"uses_additional_data":false},{"leaderboard":"/sota/multi-label-classification-on-mlrsnet","task":"Multi-Label Classification","dataset":"MLRSNet","model":"ResNet50 (scratch)","rank_in_archive_order":2,"of":2,"metrics":{"F1-score":"91.83"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}