Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
FSDETR, a frequency-spatial feature enhancement framework, is introduced to improve small object detection by addressing feature degradation, occlusion, and background interference. The framework incorporates a Spatial Hierarchical Attention Block (SHAB) for enhanced semantic representation, a Deformable Attention-based Intra-scale Feature Interaction (DA-AIFI) to mitigate occlusion, and a Frequency-Spatial Feature Pyramid Network (FSFPN) to preserve fine-grained details. Experiments demonstrate that FSDETR achieves state-of-the-art performance on VisDrone 2019 and TinyPerson benchmarks with only 14.7M parameters.
Achieve state-of-the-art small object detection with a 14.7M parameter model by fusing frequency and spatial feature enhancement.
Small object detection remains a significant challenge due to feature degradation from downsampling, mutual occlusion in dense clusters, and complex background interference. To address these issues, this paper proposes FSDETR, a frequency-spatial feature enhancement framework built upon the RT-DETR baseline. By establishing a collaborative modeling mechanism, the method effectively leverages complementary structural information. Specifically, a Spatial Hierarchical Attention Block (SHAB) captures both local details and global dependencies to strengthen semantic representation. Furthermore, to mitigate occlusion in dense scenes, the Deformable Attention-based Intra-scale Feature Interaction (DA-AIFI) focuses on informative regions via dynamic sampling. Finally, the Frequency-Spatial Feature Pyramid Network (FSFPN) integrates frequency filtering with spatial edge extraction via the Cross-domain Frequency-Spatial Block (CFSB) to preserve fine-grained details. Experimental results show that with only 14.7M parameters, FSDETR achieves 13.9% APS on VisDrone 2019 and 48.95% AP50 tiny on TinyPerson, showing strong performance on small-object benchmarks. The code and models are available at https://github.com/YT3DVision/FSDETR.