Structure-Preserving Super Resolution with Gradient Guidance
这是20年3月的一篇cvpr。 该文章提出了现有的基于对抗思想的超分辨率重建网络虽然能够很好的恢复出逼近真实世界的纹理细节，但恢复出的图像常常会伴随这图像失真的问题，因此这篇文章将梯度图引入对抗网络中指导图像重建消除图像失真的问题。 网络架构方面采用了两条支路构成网络，一条重建网络，一条生成梯度图的网络，利用lr图像生成lr的梯度图，此处M(·)与现有生成梯度图的方法一致，相邻行与列元素相减后进行
相关文章
- 1. Structure-Preserving Super Resolution with Gradient Guidance 译文SPSR
- 2. 《Structure-Preserving Super Resolution with Gradient Guidance》论文阅读
- 3. Enhancing Perceptual Loss with Adversarial Feature Matching for Super-Resolution
- 4. SRFeat: Single Image Super-Resolution with Feature Discrimination
- 5. Unified Dynamic Convolutional Network for Super-Resolution with Variational Degradations
- 6. Lightweight Image Super-Resolution with Adaptive Weighted Learning Network
- 7. 2020ECCV VSR Video Super-Resolution with Recurrent Structure-Detail Network
- 8. Audio Bit Depth Super-Resolution with Neural Networks
- 9. Robust Video Super-Resolution with Learned Temporal Dynamics
- 10. 超分(Super-Resolution, SR )
- 更多相关文章...
- • XSLT
- • Kotlin 继承 - Kotlin 教程
- • 为了进字节跳动，我精选了29道Java经典算法题，带详细讲解
- • RxJava操作符（四）Combining
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. Structure-Preserving Super Resolution with Gradient Guidance 译文SPSR
- 2. 《Structure-Preserving Super Resolution with Gradient Guidance》论文阅读
- 3. Enhancing Perceptual Loss with Adversarial Feature Matching for Super-Resolution
- 4. SRFeat: Single Image Super-Resolution with Feature Discrimination
- 5. Unified Dynamic Convolutional Network for Super-Resolution with Variational Degradations
- 6. Lightweight Image Super-Resolution with Adaptive Weighted Learning Network
- 7. 2020ECCV VSR Video Super-Resolution with Recurrent Structure-Detail Network
- 8. Audio Bit Depth Super-Resolution with Neural Networks
- 9. Robust Video Super-Resolution with Learned Temporal Dynamics
- 10. 超分(Super-Resolution, SR )