FSDC-DETR: A Frequency-Spatial Domain Collaborative DETR for Small Object Detection

2026-07-06

AI Summary

This work addresses the challenge of small object detection, which suffers from high-frequency information loss due to spatial aggregation and spectral aliasing in the frequency domain. To mitigate these issues, we propose a novel frequency–spatial collaborative detection Transformer framework that explicitly models the complementary representations of both domains for the first time. Our approach integrates three key components: Dual-Branch Feature Self-Adaptive Fusion (DBFSAF), Split Frequency–Spatial Feature Fusion (SFS-FF), and Frequency–Spatial Dynamic Downsampling (FSD-Down), collectively preserving high-frequency details and enabling progressive cross-scale interactions. Extensive experiments demonstrate significant performance gains, with absolute AP improvements of 6.4 and 6.6 on VisDrone-DET2019 and AITODv2, respectively, and notable gains of 6.8 and 6.9 in small-object AP, establishing new state-of-the-art results.

Abstract

Small object detection (SOD) remains a challenging task in real-world applications. Despite recent advances, existing detectors remain limited by rigid processing that entangle spatial aggregation with implicit frequency aliasing and truncation, leading to inadequate preservation of high-frequency components for SOD. To tackle these limitations, we propose a Frequency-Spatial Domain Collaborative Detection Transformer (FSDC-DETR), a novel collaborative framework that explicitly models complementary spatial and frequency representations. Specifically, we first introduce Dual-Branch Frequency-Spatial Adaptive Fusion (DBFSAF) to enhance frequency diversity and adaptively capture frequency-spatial domain discriminative representations. Building on these representations, a frequency-spatial interaction scheme is further explored within the hybrid encoder to enable progressive feature propagation to the decoder. In particular, structure-aware frequency-spatial aggregation is achieved through Shunt Frequency-Spatial Feature Fusion (SFS-FF), establishing bidirectional interaction and progressive cross-scale propagation between frequency and spatial representations for coherent discriminative modeling. Meanwhile, informative high-frequency responses are preserved during scale transitions through Frequency-Spatial Dynamic Downsampling (FSD-Down), thereby minimizing frequency degradation throughout multi-scale fusion for the precise SOD. Experimental results demonstrate that FSDC-DETR achieves state-of-the-art performance, improving AP by 6.4 on VisDrone-DET2019 and 6.6 on AITODv2, with gains of 6.8 and 6.9 AP for small objects. The code is available at github.com/nevereverinsomnia/FSDC-DETR.

Problem

• Small Object Detection
• Frequency Domain
• Spatial Domain
• High-Frequency Preservation
• Object Detection

Innovation

• Frequency-Spatial Collaboration
• Small Object Detection
• Adaptive Fusion
• Dynamic Downsampling
• DETR