【论文翻译】Distort-and-Recover: Color Enhancement using Deep Reinforcement Learning
ABSTRACT： 本文提出——人类修图是按照步骤逐渐进行的序列，于是用MDP建模，训练agent得到一个最优的动作序列。此外，我们提出了一种“失真-复原”训练方案，只需要高质量图像进行训练，而不需要“修改前后的图像对“。资料和代码在https://sites.google.com/view/distort-and-recover/可以找到。 论文地址 https://arxiv.org/pdf/
相关文章
- 1. 【论文翻译】Playing Atari with Deep Reinforcement Learning
- 2. 论文翻译：Deep Progressive Reinforcement Learning for Skeleton-based Action Recognition
- 3. 【论文翻译】Deep Residual Learning for Image Recognition
- 4. 暗光增强论文：Underexposed Photo Enhancement using Deep Illumination Estimation翻译与解读
- 5. 【论文翻译】Deep residual learning for image recognition
- 6. [论文阅读]DeepOtsu：Document Enhancement and Binarization using Iterative Deep Learning
- 7. 论文翻译：2D/3D Pose Estimation and Action Recognition using Multitask Deep Learning
- 8. 论文阅读笔记（五十二）：Outline Objects using Deep Reinforcement Learning
- 9. 【论文翻译】Deep Learning for Classical Japanese Literature
- 10. Deep Reinforcement Learning
- 更多相关文章...
- • Eclipse 编译项目 - Eclipse 教程
- • XSL-FO color-profile 对象 - XSL-FO 教程
- • Scala 中文乱码解决
- • 三篇文章了解 TiDB 技术内幕——说存储
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. 【论文翻译】Playing Atari with Deep Reinforcement Learning
- 2. 论文翻译：Deep Progressive Reinforcement Learning for Skeleton-based Action Recognition
- 3. 【论文翻译】Deep Residual Learning for Image Recognition
- 4. 暗光增强论文：Underexposed Photo Enhancement using Deep Illumination Estimation翻译与解读
- 5. 【论文翻译】Deep residual learning for image recognition
- 6. [论文阅读]DeepOtsu：Document Enhancement and Binarization using Iterative Deep Learning
- 7. 论文翻译：2D/3D Pose Estimation and Action Recognition using Multitask Deep Learning
- 8. 论文阅读笔记（五十二）：Outline Objects using Deep Reinforcement Learning
- 9. 【论文翻译】Deep Learning for Classical Japanese Literature
- 10. Deep Reinforcement Learning