随笔分类 - Dehaze papers
摘要:题目:Rethinking Performance Gains in Image Dehazing Networks paper: https://arxiv.org/pdf/2209.11448v1.pdf code: https://github.com/idkiro/gunet 摘要:图像去雾
阅读全文
摘要:题目:Compound Multi-branch Feature Fusion for Real Image Restoration paper:https://arxiv.org/pdf/2206.02748v1.pdf code:https://github.com/FanChiMao/CMFN
阅读全文
摘要:Deep-Energy(2021-IEEE) Unsupervised single image dehazing with generative adversarial network 摘要 最新的单幅图像去模糊学习算法被设计为使用成对的模糊图像和相应的地面真实图像(通常是合成图像)进行训练。真正
阅读全文
摘要:UCL-Dehaze(2022-CVPR) UCL-Dehaze: Towards Real-world Image Dehazing via Unsupervised Contrastive Learning paper: https://arxiv.org/abs/2205.01871 code
阅读全文
摘要:FD-GAN(2018-NeurIPS) FD-GAN: Generative Adversarial Networks with Fusion-discriminator for Single Image Dehazing 摘要 近年来,卷积神经网络(CNN)在单幅图像去噪方面取得了很大的进步,在
阅读全文
摘要:KCKE(2022-CVPR) Learning Multiple Adverse Weather Removal via Two-stage Knowledge Learning and Multi-contrastive Regularization: Toward a Unified Mode
阅读全文
摘要:D4(2022-CVPR) D4: Self-augmented Unpaired Image Dehazing via Density and Depth Decomposition 摘要 为了克服在合成模糊-干净图像对上训练的去雾模型的过度拟合问题,最近的许多方法试图通过对非配对数据进行训练来提
阅读全文
摘要:KDDN(2020-CVPR) Distilling Image Dehazing with Heterogeneous Task Imitation 摘要 最先进的深度除雾模型在训练中往往是困难的。知识蒸馏使得利用教师网络训练学生网络成为可能。然而,大多数的知识蒸馏方法都是应用于图像分类、语义分割
阅读全文
摘要:FFA-Net (2020-AAAI) FFA-Net Feature Fusion Attention Network for Single Image Dehazing 摘要 在本文中,我们提出了一种端到端的特征融合注意力网络(FFA-Net)来直接恢复无雾图像。 FFA-Net 架构由三个关键
阅读全文