FDEP-Framework

KO | EN

gitlite — search

#typescript #ai-agents #deepseek-harness #dsh-plugin #open-source #ai #cli #dsh #claude-code #codex #developer-tools #react

★ 43 Open GitHub ↗

Official code of the paper "Rethinking Infrared Small Target Detection: A Foundation- Driven Efficient Paradigm"

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<div align="center">
 <h1 style="border-bottom: none;">Rethinking Infrared Small Target Detection: A Foundation-Driven Efficient Paradigm</h1>
</div>

<div align="center">
 
 <a href="https://scholar.google.com/citations?user=Dd4_VW8AAAAJ&hl=zh-CN" target="_blank">Chuang Yu</a>1,2,6,&nbsp
 <a href="https://scholar.google.com/citations?user=3cBa6r4AAAAJ&hl=zh-CN" target="_blank">Jinmiao Zhao</a>1,2,&nbsp
 <a>Yunpeng Liu</a>1*,&nbsp
 <a href="https://iron-lyk.github.io/" target="_blank">Yaokun Li</a>3,&nbsp
 <a>Xiujun Shu</a>4,
 
 <a>Yuanhao Feng</a>4,&nbsp
 <a>Bo Wang</a>4,&nbsp
 <a href="https://scholar.google.com/citations?user=y5Ov6VAAAAAJ&hl=zh-CN" target="_blank">Yimian Dai</a>5,&nbsp
 <a href="https://scholar.google.com/citations?user=-xQ-C1sAAAAJ&hl=zh-CN" target="_blank">Xiangyu Yue</a>6*
 
 
 1 Shenyang Institute of Automation, Chinese Academy of Sciences
 
 2 University of Chinese Academy of Sciences &nbsp;&nbsp;
 
 3 Sun Yat-sen University &nbsp;&nbsp;
 4 Tencent &nbsp;&nbsp;
 5 Nankai University &nbsp;&nbsp;
 6 MMLab, The Chinese University of Hong Kong
 
</div>

## 💥 Abstract
While large-scale visual foundation models (VFMs) exhibit strong generalization across diverse visual domains, their potential for single-frame infrared small target (SIRST) detection remains largely unexplored. **To fill this gap, we systematically introduce the frozen representations from VFMs into the SIRST task for the first time and propose a Foundation-Driven Efficient Paradigm (FDEP), which can seamlessly adapt to existing encoder-decoder-based methods and significantly improve accuracy without additional inference overhead.** Specifically, a Semantic Alignment Modulation Fusion (SAMF) module is designed to achieve dynamic alignment and deep fusion of the global semantic priors from VFMs with task-specific features. Meanwhile, to avoid the inference time burden introduced by VFMs, we propose a Collaborative Optimization-based Implicit Self-Distillation (CO-ISD) strategy, which enables implicit semantic transfer between the main and lightweight branches through parameter sharing and synchronized backpropagation. In addition, to unify the fragmented evaluation system, we construct a Holistic SIRST Evaluation (HSE) metric that performs multi-threshold integral evaluation at both pixel-level confidence and target-level robustness, providing a stable and comprehensive basis for fair model comparison. Extensive experiments demonstrate that the SIRST detection networks equipped with our FDEP framework achieve state-of-the-art (SOTA) performance on multiple public datasets.

## 🚀 Overview
<div align="center">
 
 <img width="100%" src="imgs/FDEP-Framework-Overview.png" alt="FDEP-Framework-Overview">
 
</div>

## ✅ TODO List
We are finalizing the release of the paper, dataset and code and aim to complete it as soon as possible. Please stay tuned! ⚡⚡⚡
- [x] Release paper. [[Paper/ArXiv](https://arxiv.org/abs/2512.05511)] 
- [ ] Release training and inference code.

## Other link 
1. My homepage: [[YuChuang](https://github.com/YuChuang1205)]
2. My "PAL Framework" project code (ICCV2025)：[[Link](https://github.com/YuChuang1205/PAL)]

FDEP-Framework

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README