{"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/scinet-spatial-and-contrast-interactive-super","title":"SCINet: Spatial and Contrast Interactive Super-Resolution Assisted Infrared UAV Target Detection","arxiv_id":null,"date":"2024-10-01","proceeding":"IEEE Transactions on Geoscience and Remote Sensing 2024 10","authors":["Houzhang Fang","Lan Ding","Xiaolin Wang","Yi Chang","Luxin Yan","Li Liu","Jinrui Fang"],"abstract":"Unmanned aerial vehicle (UAV) detection based on\r\nthermal infrared imaging has been one of the most important\r\nsensing technologies in the anti-UAV system. However, the\r\ntechnical limitations and long-range detection of thermal sensors\r\noften lead to acquiring low-resolution (LR) infrared images,\r\nthereby bringing great challenges for the subsequent target\r\ndetection task. In this article, we propose a novel spatial\r\nand contrast interactive super-resolution network (SCINet) for\r\nassisting infrared UAV target detection. The network consists\r\nof two main subnetworks: a spatial enhancement branch (SEB)\r\nand a contrast enhancement branch (CEB). The SEB embeds\r\nthe lightweight convolution module and attention mechanism\r\nto highlight the spatial structure detail features of infrared\r\nUAV targets. The proposed CEB incorporates the centeroriented\r\ncontrast-aware module and multibranch collapsible\r\nmodule, which can provide local contrast priors to reconstruct\r\nthe super-resolved UAV target image. The spatial features\r\nof the intermediate layers from the SEB are integrated\r\ninto the CEB as a rich gradient prior. Besides, the output\r\nfeatures of the CEB are aggregated into those of the SEB\r\nfor further supplementing the contrast of spatial features in\r\nreturn. Dual-branch feature interaction (DBFI) of the SEB\r\nand CEB can further enhance the spatial details and target\r\nsaliency of the targets. In addition, we also introduce an\r\ninfrared UAV detection network via a new dual-dimensional\r\nfeature calibration module (DFCM) for boosting the detection\r\nperformance. Extensive experiments demonstrate that the SCINet\r\noutperforms the state-of-the-art (SOTA) SR methods on real\r\ninfrared UAV sequences and improves the detection performance\r\nof infrared small UAV targets. The code is available at\r\nhttps://github.com/IVPLaboratory/SCINet.","url_abs":"https://ieeexplore.ieee.org/document/10701558","url_pdf":"https://ieeexplore.ieee.org/document/10701558","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":"scinet-spatial-and-contrast-interactive-super","repo_url":"https://github.com/IVPLaboratory/SCINet","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"scinet-spatial-and-contrast-interactive-super","repo_url":"https://github.com/MindSpore-scientific/code-14/tree/main/SciNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}