From Spatial to Spectral: An Efficient, Frequency-Guided Feature Representation Learner for Small Object Detection

2026-06-22

AI Summary

This work addresses the challenges of small object detection, which suffers from feature sparsity and the loss of high-frequency details inherent in spatial-domain approaches. To overcome these limitations, the authors propose a novel spectral-domain feature learning paradigm featuring a lightweight, plug-and-play Decomposition-Enhancement-Reconstruction (DER) operator. This operator integrates frequency-aware modulation into the backbone, neck, and detection head through three key components: a Wavelet Difference Gate (WDG), a Log-Gabor Enhancer (LGE), and a Frequency-Domain-driven Detection Head (FDHead). The resulting framework is architecture-agnostic, compatible with both CNNs and Transformers, and achieves state-of-the-art performance on benchmarks such as VisDrone2019 and UAVDT—significantly outperforming YOLOv11 at one-sixth the parameter count while enabling more precise localization of small objects.

Abstract

Efficient small object detection is bottlenecked by the inherent feature scarcity of tiny targets, which is further aggravated by operations of spatial-domain detectors that indiscriminately discard critical high-frequency details. Recovering these fragile cues within the spatial domain is notoriously difficult, as it often requires computationally expensive architectural upscaling that inadvertently amplifies background noise. To bridge this gap, we propose a paradigm \textbf{shift from spatial to spectral} feature processing, introducing a holistic solution with the following novelty: (1) A versatile \textbf{Frequency-Guided Feature Representation framework} that generalizes across diverse detector architectures (both CNN and Transformer-based), offering a robust alternative to spatial-only feature extraction; (2) The unified \textbf{Decompose--Enhance--Reconstruct (DER)} operator, instantiated via three \textbf{lightweight, plug-and-play} modules -- Wavelet-Difference Gate (WDG), Log-Gabor Enhancer (LGE), and Frequency-Driven Head (FDHead) -- to systematically inject frequency-aware modulation into the backbone, neck, and head. This mechanism decouples feature modeling from resolution reduction, capturing discriminative high-frequency components to enable accurate localization with significantly reduced parameter redundancy; (3) Extensive validation on multi-domain benchmarks (VisDrone2019, UAVDT, TinyPerson, DOTAv1) demonstrating consistent gains. Notably, our proposed \textbf{DERNet} series outperforms YOLOv11 models under the same scale while requiring \textbf{only 1/6 of the parameters}, backed by rigorous spectral diagnostics and error decomposition analysis.

Problem

• small object detection
• feature scarcity
• high-frequency details
• spatial-domain processing
• computational efficiency

Innovation

• Frequency-Guided Feature Representation
• Decompose–Enhance–Reconstruct (DER)
• Small Object Detection
• Spectral Domain Processing
• Lightweight Plug-and-Play Modules