{"type":"ScholarlyArticle","title":"Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection","authors":[{"name":"Yanchao Bi","slug":"yanchao-bi","given":"Yanchao","family":"Bi","affiliation":""},{"name":"Yang Ning","slug":"yang-ning","given":"Yang","family":"Ning","affiliation":""},{"name":"Xiushan Nie","slug":"xiushan-nie","given":"Xiushan","family":"Nie","affiliation":""},{"name":"Xiankai Lu","slug":"xiankai-lu","given":"Xiankai","family":"Lu","affiliation":""},{"name":"Yongshun Gong","slug":"yongshun-gong","given":"Yongshun","family":"Gong","affiliation":""},{"name":"Leida Li","slug":"leida-li","given":"Leida","family":"Li","affiliation":""}],"abstract":"Current feature fusion strategies often fail to adequately account for the influence of activation intensity across different scales on small object features, which impedes the effective detection of small objects. To address this limitation, we propose the Region-Adaptive Feature Disentanglement and Enhancement (RAFDE) strategy, which improves both downsampling and feature fusion by leveraging activation intensity variations at multiple scales. First, we introduce the Boundary Transitional Region-enhanced Downsampling (BTRD) module, which enhances boundary transitional regions containing both strongly and weakly activated features, thereby mitigating the loss of crucial boundary information for small objects. Second, we present the Regional-Adaptive Feature Fusion (RAFF) module, which adaptively disentangles and fuses co-activated and uni-activated regions from adjacent levels into the current level, effectively reducing the risk of small objects being overwhelmed. Extensive experiments on several public datasets demonstrate that the RAFDE strategy is highly effective and outperforms state-of-the-art methods. The code is available at https://github.com/b-yanchao/RAFDE.git.","doi":"10.24963/ijcai.2025/78","url":"https://ijcai.org/proceedings/2025/78","pdf":"https://ijcai.org/proceedings/2025/0078.pdf","markdown":"https://ijcai.org/proceedings/2025/78.md","bibtex":"https://ijcai.org/proceedings/2025/bibtex/78","volume":{"slug":"2025","title":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2025","url":"https://ijcai.org/proceedings/2025/","year":2025,"location":"Montreal, Canada","dates":"16-22 August 2025","editors":"James Kwok","isbn":"978-1-956792-06-5","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2025"},"section":"Computer Vision","track":"Main Track","number":78,"pages":{"first":693,"last":701},"keywords":["Machine Learning: ML: Classification","Machine Learning: ML: Convolutional networks","Computer Vision: CV: Recognition (object detection, categorization)","Computer Vision: CV: Machine learning for vision"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}