高光谱图像修复笔记

目录

RetinexFormer

也有MST-plus-plus代码，分辨率可以调

MST-plus-plus

github地址：

WACV2023 DSTrans

---

RetinexFormer

GitHub - caiyuanhao1998/Retinexformer: "Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement" (ICCV 2023) & (NTIRE 2024 Challenge)

有网络结构：

basicsr/models/archs/RetinexFormer_arch.py

RetinexFormer不能自由修改通道

也有MST-plus-plus代码，分辨率可以调

if __name__ == '__main__':# from fvcore.nn import FlopCountAnalysisdevice = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")model = MST_Plus_Plus(in_channels=2,out_channels=4).to(device)inputs = torch.randn((1, 2, 256, 256)).to(device)out=model(inputs)print(out.shape)

MST-plus-plus

CVPR录用+NTIRE冠军！清华提出首个高光谱图像重建Transformer__财经头条

MST++ 是 MST 的后续工作，其全称为 Multi-stage Spectral-wise Transformer。

github地址：

GitHub - caiyuanhao1998/MST-plus-plus: "MST++: Multi-stage Spectral-wise Transformer for Efficient Spectral Reconstruction" (CVPRW 2022) & (Winner of NTIRE 2022 Spectral Recovery Challenge) and a toolbox for spectral reconstruction

WACV2023 DSTrans

Dual-Stream Transformer for Hyperspectral Image Restoration

[CVPR 2022 & NTIRE 冠军] 首个高光谱图像重建Transformer - 知乎