Adaptive Spectrum-Aware Feature Disentangled Network for Small Object Detection

2026-06-27

AI Summary

Small object detection is hindered by low resolution and weak visual cues, rendering it highly susceptible to background interference in the spectral domain. To address this challenge, this work proposes SFDNet, a novel framework that introduces an Adaptive Spectral Decoupling (ASD) module to decompose backbone features into multiple complementary spectral components, effectively suppressing background noise. Additionally, a Class Prototype Distillation (CPD) mechanism is devised to enhance semantic consistency and intra-class compactness among small objects of the same category through knowledge distillation. Extensive experiments on multiple established small object detection benchmarks demonstrate that SFDNet significantly outperforms current state-of-the-art methods, confirming its effectiveness and robustness.

Abstract

Small Object Detection (SOD) is a fundamental yet challenging problem in computer vision due to its limited spatial resolution and weak visual cues. Although recent approaches have achieved remarkable advances, the background distractors in different frequency spectra still degrade the performance. In this paper, we propose a novel small object detection framework termed SFDNet, which is capable of detecting small objects via efficient spectrum-aware feature disentanglement. Specifically, we propose an Adaptive Spectrum Disentanglement (ASD) module that decomposes backbone features into multiple complementary spectral components, aiming to construct discriminative object-relevant representations by discarding the background distractors for each component. Afterwards, to strengthen the semantic consistency of the similar objects in the same class, we propose a Class-Wise Prototype Distillation (CPD) procedure, which establishes class prototypes for the object instances and enforces the compact representation by efficient prototype distillation. Extensive experiments on multiple challenging benchmarks show that SFDNet outperforms existing state-of-the-art methods by a large margin. Our code is available at https://github.com/ManOfStory/SFDNet.

Problem

• Small Object Detection
• Background Distractors
• Frequency Spectra
• Feature Disentanglement
• Visual Cues

Innovation

• Adaptive Spectrum Disentanglement
• Feature Disentanglement
• Small Object Detection
• Prototype Distillation
• Spectrum-Aware Representation