PPIFuse (IEEE TCSVT 2025)

---

This is the official implementation of the PPIFuse model proposed in the paper (PPIFuse: Physical Priors Injected Infrared and Visible Image Fusion) with pytorch.

---

Network Architecture

---

Requirements

---

• python 3.8
• torch 2.4.1+cu118
• torchaudio 2.4.1+cu118
• torchvision 0.19.1+cu118
• scipy 1.10.1
• numpy 1.24.1
• tqdm 4.67.1
• matplotlib 3.7.5

Training

---

Please run

python run_training_PPIFuse.py

Pre-trained models & Testing

---

Download pre-trained models from here, password pihh. Then run

python run_test.py.

Fusion results

---

The fusion results for the TNO, MSRS, M3FD, and LLVIP datasets are available here, password n2py.

Citation

---

If you have any questions, feel free to contact us(zhangqh202503@163.com or liuqiao.hit@gmail.com).