Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection

Bi, Yanchao; Ning, Yang; Nie, Xiushan; Lu, Xiankai; Gong, Yongshun; Li, Leida

doi:10.24963/ijcai.2025/78

Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection

Yanchao Bi, Yang Ning, Xiushan Nie, Xiankai Lu, Yongshun Gong, Leida Li

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 693-701.

https://doi.org/10.24963/ijcai.2025/78

Keywords

Machine Learning: ML: Classification
Machine Learning: ML: Convolutional networks
Computer Vision: CV: Recognition (object detection, categorization)
Computer Vision: CV: Machine learning for vision

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.

Yanchao Bi, Yang Ning, Xiushan Nie, Xiankai Lu, Yongshun Gong, and Leida Li. 2025. Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 693–701. https://doi.org/10.24963/ijcai.2025/78

@inproceedings{ijcai2025p78,
  title     = {Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection},
  author    = {Bi, Yanchao and Ning, Yang and Nie, Xiushan and Lu, Xiankai and Gong, Yongshun and Li, Leida},
  booktitle = {Proceedings of the Thirty-Fourth International Joint Conference on
               Artificial Intelligence, {IJCAI-25}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {James Kwok},
  pages     = {693--701},
  year      = {2025},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2025/78},
  url       = {https://doi.org/10.24963/ijcai.2025/78},
}

TY  - CPAPER
TI  - Towards Region-Adaptive Feature Disentanglement and Enhancement for Small Object Detection
AU  - Bi, Yanchao
AU  - Ning, Yang
AU  - Nie, Xiushan
AU  - Lu, Xiankai
AU  - Gong, Yongshun
AU  - Li, Leida
T2  - Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2025
SP  - 693
EP  - 701
DO  - 10.24963/ijcai.2025/78
UR  - https://doi.org/10.24963/ijcai.2025/78
ER  -