Image Inpainting2020——图像修复 —— 论文汇总
【摘要】 Image Inpainting2020——图像修复 —— 论文汇总
🥇 版权: 本文由【墨理学AI】原创、各位大佬、一文读懂、敬请查阅 🎉 声明: 作为全网 AI 领域 干货最多的博主之一，不负光阴不负卿
📔 种草一个 论文总结网站
备注：感觉自己总结的，啥也不是…
- paperswithcode
- Awesome-Image-Inpainting–论文汇总
- DrogoZhang 博主归纳的图像修复(Image Inpainting)必读论文——2000–2019
- https://www.zhihu.com/question/56801298| 深度学习做图像修复
| Paper Name | Architecture | Inpaint type | time |
|---|---|---|---|
| Context Encoders: Feature Learning by Inpainting | Context Encoders + treconstruction and adversarial loss （ Semantics 修复 ） | Random block | 2016 |
| Globally and Locally Consistent Image Completion | Completion Network + Global and Local Discriminator | 不规则洞 、目标移除 | 2017 |
| ---- | CVPR | – | – |
| Image Inpainting for Irregular Holes Using Partial Convolutions | encoder-decoder + Partial Convolutions | 不规则洞 | 2018 |
| Free-Form Image Inpainting with Gated Convolution | Coarse(粗糙的) Network + Contextual Attention or ( gated convolution and SN-PatchGAN) | 目标移除、图像（人脸）编辑 | 2019 |
| 基于动差重构损失的模范生成对抗网络（ExMRGAN） | 动差重构损失 GAN | 眼部遮挡修复 | 2019 |
| Learning Pyramid-Context Encoder Network for High-Quality Image Inpainting | encoder-decoder + Cross-layer attention transfer + Pyramid filling | 块遮挡、不规则洞 | 2019 |
| EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning | edge generation and image completion gan DilatedConv+ResidualBlocks | 块遮挡、不规则洞 | 2019 |
| 3D Photography using Context-aware Layered Depth Inpainting | edge inpainting network （GAN）+ Context-aware color and depth inpainting (U-Net) | single RGB-D input image into a 3D photo | 2020 |
| Prior Guided GAN Based Semantic Inpainting | noise prior（LSTM-CNN） + GAN ( 先训练GAN-再训练先验) | 块遮挡 | 2020 |
| Recurrent Feature Reasoning for Image Inpainting | area identification module + feature reasoning module + feature merging operator PConv+Conv+ KCA(Knowledge Consistent Attention) | 不规则洞 | 2020 |
| UCTGAN: Diverse Image Inpainting based on Unsupervised Cross-Space Translation | manifold projection & conditional encoder + generation module 、KL Divergence Loss、Cross Semantic Attention | 块遮挡 | 2020 |
| Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting | Contextual Residual Aggregation GAN + attention score sharing、 multi-scale attention transfer、 Light-Weight Gated Convolutions (LWGC) 、 dilated convolutions | 目标移除、不规则洞 | 2020 |
| Assessing Eye Aesthetics for Automatic Multi-Reference Eye In-Painting | GAN + eye aesthetic(Encoder 、Residua Block、Decoder) and face semantic (目标轮廓检测网络) | 眼部修复 | 2020 |
| Bringing Old Photos Back to Life | Two VAE(real photos 、 clean images) jointly training an adversarial discriminator + Restoration through latent mapping | 照片划痕 | 2020 |
| – | ECCV | – | – |
| Rethinking Image Inpainting via a Mutual Encoder-Decoder with Feature Equalizations | channel and spatial domains + Partial Conv | 不规则洞 | 2020 |
| Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation | 判别器来作为距离度量+先整体后局部的渐进式重建 | 图像修复多种任务 | 2020 |
📕 论文对应源码
| paper | code | time | Interest index | type |
|---|---|---|---|---|
| Pluralistic Image Completion | https://github.com/lyndonzheng/Pluralistic-Inpainting | 2019 | 4 | Pytoch 0.4.0 |
| Free-Form Image Inpainting with Gated Convolution | https://github.com/JiahuiYu/generative_inpainting | 2019 | 4 | tensorflow 1.X |
| EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning | https://github.com/knazeri/edge-connect | 2019 | 4 | PyTorch 1.0 |
| Recurrent Feature Reasoning for Image Inpainting | https://github.com/jingyuanli001/RFR-Inpainting | 2020 | 5 | PyTorch >= 1.0.0 |
📗 论文解析
| paper | content | link | time |
|---|---|---|---|
| Image Inpainting for Irregular Holes Using Partial Convolutions | 用部分卷积对不规则孔进行图像修复 | https://blog.csdn.net/king_haw/article/details/86177796 | 2018 |
| Pluralistic Image Completion | 从多元化的层面对图像进行修补 | https://blog.csdn.net/sunlum/article/details/106673915 | 2019 |
| Generative Image Inpainting with Contextual Attention | Contextual Attention | https://zhuanlan.zhihu.com/p/50620348 | 2018 |
| PEPSI : Fast Image Inpainting with Parallel Decoding Network | coarse-to-fine network with the contextual attention module (CAM) + 的改进 ；编码器共享；解码器部分参数共享 – 可惜没有源码 | https://blog.csdn.net/qq_40079310/article/details/100166330 | 2019 |
📘 近3年图像修复方向 30 + 论文 获取途径如下
搜索关注本博客同名公号，公号后台，回复【图像修复2020】获取 近3年 30 + 论文下载链接（含本博文中提及论文）：
图像修复2020
📘 更多有效教程
此次博文内容难以 以偏概全，如有不恰当的地方，欢迎评论区批评指正
对于即将入行计算机视觉的小伙伴，墨理这里推荐收藏的干货博文目前如下
🚀🚀 墨理学AI
🎉 作为全网 AI 领域 干货最多的博主之一，❤️ 不负光阴不负卿 ❤️ ❤️ 如果文章对你有帮助、点赞、评论鼓励博主的每一分认真创作
【版权声明】本文为华为云社区用户原创内容，转载时必须标注文章的来源（华为云社区）、文章链接、文章作者等基本信息， 否则作者和本社区有权追究责任。如果您发现本社区中有涉嫌抄袭的内容，欢迎发送邮件进行举报，并提供相关证据，一经查实，本社区将立刻删除涉嫌侵权内容，举报邮箱：
cloudbbs@huaweicloud.com
- 点赞
- 收藏
- 关注作者
评论（0）