LCPNet: Latent Consistent Proximal Unfolding Network for Infrared Small Target Detection

2026-07-05

AI Summary

This work addresses the limitations of existing infrared small target detection methods, which often fail to effectively model the physical decomposition structure between targets and background, thereby compromising accuracy and robustness. To overcome this, the authors propose a latent-space deep unfolding network that incorporates a low-rank prior and embeds a physics-based optimization process within the latent space. Key innovations include the first-ever enforcement of low-rank constraints during optimization unfolding in latent space, the design of a consistent proximal solver with a shared optimization memory (SOM) mechanism to enhance target update stability and inter-stage collaboration, and the integration of task-adaptive normalization with gain control. Evaluated on four public datasets, the method significantly outperforms state-of-the-art approaches, achieving high detection accuracy, low false alarm rates, and favorable computational efficiency.

Abstract

Infrared small target detection (IRSTD) aims to identify long distance small targets from complex infrared backgrounds, and is a fundamental task in remote sensing. Deep learning methods have improved IRSTD by learning discriminative image-to-mask mappings, but such feed-forward designs often underuse physical decomposition structure between targets and backgrounds. Deep unfolding methods partially address this issue by embedding model-driven iterations into neural networks, yet existing designs still operate mainly in image domain and use updates and memory mechanisms that are not fully coupled with underlying optimization process. To address these limitations, we propose Latent Consistent Proximal unfolding network (LCPNet). First, we verify that low-rank prior remains valid in latent representations and perform unfolding in this space, preserving physical constraint while avoiding repeated compression of intermediate states. Second, we derive a Latent Consistent Proximal (LCP) solver that evolves each latent variable from its previous state rather than reconstructing through an indirect residual, and stabilizes small target updates through task-adaptive normalization and gain control. Third, we introduce Shared Optimization Memory (SOM), a common historical state shared by all decomposition variables to provide coordinated guidance across unfolding stages. Extensive experiments on four public benchmarks demonstrate that LCPNet outperforms state-of-the-art methods while achieving accurate and robust detection with low false alarms and competitive efficiency. Model and code are available at https://github.com/Tianfang-Zhang/LCPNet.

Problem

• Infrared Small Target Detection
• Deep Unfolding
• Physical Decomposition
• Latent Representation
• Optimization Process

Innovation

• deep unfolding
• latent space
• low-rank prior
• proximal solver
• shared optimization memory