2019年CV领域，值得一看的综述文章！_deep learning in video multi-object tracking: a su

问题：2019年，CV领域，你推荐哪些综述性的文章？

https://www.zhihu.com/question/355566860

 

知乎高质量回答

 

1、作者：Amusi

https://www.zhihu.com/question/355566860/answer/894352980

 

一直关注CV这一块，下面分享几个2019年比较好的CV综述，方向涵盖：目标检测、图像分割、目标跟踪和超分辨率等

 

• 目标检测

 

2019 四大目标检测综述论文：

 

Imbalance Problems in Object Detection: A Review

 

• intro: under review at TPAMI

• arXiv: https://arxiv.org/abs/1909.00169

 

Recent Advances in Deep Learning for Object Detection

 

• intro: From 2013 (OverFeat) to 2019 (DetNAS)

• arXiv: https://arxiv.org/abs/1908.03673

 

A Survey of Deep Learning-based Object Detection

 

• intro：From Fast R-CNN to NAS-FPN

• arXiv：https://arxiv.org/abs/1907.09408

 

Object Detection in 20 Years: A Survey

 

• intro：This work has been submitted to the IEEE TPAMI for possible publication

• arXiv：https://arxiv.org/abs/1905.05055

 

目标检测更多论文详见：

https://github.com/amusi/awesome-object-detection

 

• 图像分割

 

Machine Learning Techniques for Biomedical Image Segmentation: An Overview of Technical Aspects and Introduction to State-of-Art Applications

 

• arXiv : https://arxiv.org/abs/1911.02521

 

Deep Semantic Segmentation of Natural and Medical Images: A Review

 

• intro: 从 FCN(2014) 到 Auto-DeepLab(2019)，本综述共含179篇语义分割和医学图像分割参考文献

• arXiv: https://arxiv.org/abs/1910.07655

 

Understanding Deep Learning Techniques for Image Segmentation

 

• intro: 本综述介绍了从2013年到2019年，主流的30多种分割算法（含语义/实例分割），50多种数据集，共计224篇参考文献

• arXiv : https://arxiv.org/abs/1907.06119

 

• 目标跟踪

 

A Review of Visual Trackers and Analysis of its Application to Mobile Robot

 

• intro: 本目标跟踪综述共含185篇参考文献！从传统方法到最新的深度学习网络

• arXiv: https://arxiv.org/abs/1910.09761

 

Deep Learning in Video Multi-Object Tracking: A Survey

 

• intro: 38页目标跟踪综述，含30多种主流算法，共计174篇参考文献

• arXiv: https://arxiv.org/abs/1907.12740

 

• 超分辨率

 

A Deep Journey into Super-resolution: A survey

 

• arXiv: https://arxiv.org/abs/1904.07523

 

Deep Learning for Image Super-resolution: A Survey

 

• arXiv: https://arxiv.org/abs/1902.06068

 

2、作者：魏秀参

 https://www.zhihu.com/question/355566860/answer/896661195

 

自荐一篇“Deep Learning for Fine-Grained Image Analysis: A Survey“：

 

《超全深度学习细粒度图像分析：项目、综述、教程一网打尽》

• 链接：

  https://mp.weixin.qq.com/s/2pJt9hlUFhR6mo1ughKkiA

 

另，除文末提及的几个具体future directions

 

• Automatic Fine-Grained Models

• Fine-Grained Few-Shot Learning

• Fine-Grained Hashing

 

之外。实际上FGIA领域还有非常多新鲜好玩的问题和应用值得探索，如：

 

• 我们围绕FGIA提出的一个目前最大的新零售场景商品数据集RPC：

  https://zhuanlan.zhihu.com/p/55627416

• 在真实细粒度识别场景中不可避免的长尾分布问题：Long tailed problems

• 存在跨域差异（Domain adaptation）的细粒度图像识别和检索

• ……