从20世纪70年代开始,有关图像检索的研究就已开始,当时主要是基于文本的图像检索技术(Text-based Image Retrieval,简称TBIR),利用文本描述的方式描述图像的特征,如绘画作品的作者、年代、流派、尺寸等。到90年代以后,出现了对图像的内容语义,如图像的颜色、纹理、布局等进行分析和检索的图像检索技术,即基于内容的图像检索(Content-based Image Retrieval,简称CBIR)技术。CBIR属于基于内容检索(Content-based Retrieval,简称CBR)的一种,CBR中还包括对动态视频、音频等其它形式多媒体信息的检索技术。

图像检索(Image Retrieval)专知荟萃

入门学习

  1. 相似图片搜索的原理 阮一峰
  2. Google 图片搜索的原理是什么?
  3. 基于内容的图像检索技(CBIR)术相术介绍
  4. 图像检索:基于内容的图像检索技术
  5. 基于内容的图像检索技术
  6. 图像检索:CNN卷积神经网络与实战 CNN for Image Retrieval
  7. 用Python和OpenCV创建一个图片搜索引擎的完整指南

综述

  1. Recent Advance in Content-based Image Retrieval: A Literature Survey. Wengang Zhou, Houqiang Li, and Qi Tian 2017
  2. Intelligent Image Retrieval Techniques: A Survey 2014
  3. A survey on content based image retrieval. 2013

进阶文章

2011

  1. Using Very Deep Autoencoders for Content-Based Image Retrieval

2013

  1. Learning High-level Image Representation for Image Retrieval via Multi-Task DNN using Clickthrough Data

2014

  1. Neural Codes for Image Retrieval
  2. Efficient On-the-fly Category Retrieval using ConvNets and GPUs

2015

  1. Learning visual similarity for product design with convolutional neural networks SIGGRAPH 2015
  2. Exploiting Local Features from Deep Networks for Image Retrieval
  3. Cross-domain Image Retrieval with a Dual Attribute-aware Ranking Network ICCV 2015
  4. Where to Buy It: Matching Street Clothing Photos in Online Shops ICCV 2015
  5. Aggregating Deep Convolutional Features for Image Retrieval
  6. Particular object retrieval with integral max-pooling of CNN activations

2016

  1. Deep Image Retrieval: Learning global representations for image search ECCV 2016
  2. Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks. CVPR 2016
  3. Fast Training of Triplet-based Deep Binary Embedding Networks. CVPR 2016
  4. Deep Relative Distance Learning: Tell the Difference Between Similar Vehicles. CVPR 2016
  5. Bags of Local Convolutional Features for Scalable Instance Search. Best Poster Award at ICMR 2016.
  6. Group Invariant Deep Representations for Image Instance Retrieval
  7. Natural Language Object Retrieval
  8. Faster R-CNN Features for Instance Search
  9. Where to Focus: Query Adaptive Matching for Instance Retrieval Using Convolutional Feature Maps
  10. Adversarial Training For Sketch Retrieval
  11. DeepFashion: Powering Robust Clothes Recognition and Retrieval with Rich Annotations
  12. CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examples
  13. PicHunt: Social Media Image Retrieval for Improved Law Enforcement
  14. The Sketchy Database: Learning to Retrieve Badly Drawn Bunnies
  15. End-to-end Learning of Deep Visual Representations for Image Retrieval
  16. What Is the Best Practice for CNNs Applied to Visual Instance Retrieval?

2017

  1. AMC: Attention guided Multi-modal Correlation Learning for Image Search. CVPR 2017
  2. Deep image representations using caption generators. ICME 2017
  3. One-Shot Fine-Grained Instance Retrieval. ACM MM 2017
  4. Selective Deep Convolutional Features for Image Retrieval. ACM MM 2017
  5. Deep Binaries: Encoding Semantic-Rich Cues for Efficient Textual-Visual Cross Retrieval. ICCV 2017
  6. Image2song: Song Retrieval via Bridging Image Content and Lyric Words. ICCV 2017
  7. SIFT Meets CNN: A Decade Survey of Instance Retrieval
  8. Image Retrieval with Deep Local Features and Attention-based Keypoints

Tutorial

  1. CVPR’16 Tutorial on Image Tag Assignment, Refinement and Retrieval
  2. Content-based image retrieval tutorial by Joani Mitro
  3. Tutorial on Image Retrieval System, (IRS)

视频教程

  1. Deep Image Retrieval: Learning global representations for image search
  2. Image Instance Retrieval: Overview of state-of-the-art

代码

  1. Neural Codes for Image Retrieval
  2. Natural Language Object Retrieval
  3. Bags of Local Convolutional Features for Scalable Instance Search
  4. Faster R-CNN Features for Instance Search
  5. CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examples
  6. Class-Weighted Convolutional Features for Visual Instance Search

领域专家

  1. Hervé Jégou
  2. Andrew Zisserman
  3. Qi Tian
  4. Artem Babenko

Datasets

  1. Corel 1000 and 10,000 图像数据库
  2. The COREL Database for Content based Image Retrieval
  3. Corel-5K and Corel -10K Datasets该页面下面给出了图片的链接,可以用python写个脚本把它们爬下来。
  4. INSTRE,中科院计算所弄的一个数据库28543张图片,还有他们做的web检索系统ISIA。
  5. MIRFLICKR 1M数据库,100多g.
  6. Image Similarity Triplet Dataset
  7. INRIA Holidays 该数据集是Herve Jegou研究所经常度假时拍的图片(风景为主),一共1491张图,500张query(一张图一个group)和对应着991张相关图像,已提取了128维的SIFT点4455091个,visual dictionaries来自Flickr60K.
  8. Oxford Buildings Dataset,5k Dataset images,有5062张图片,是牛津大学VGG小组公布的,在基于词汇树做检索的论文里面,这个数据库出现的频率极高。
  9. Oxford Paris,The Paris Dataset,oxford的VGG组从Flickr搜集了6412张巴黎旅游图片,包括Eiffel Tower等。
  10. 201Books and CTurin180 The CTurin180 and 201Books Data Sets,2011.5,Telecom Italia提供于Compact Descriptors for Visual Search,该数据集包括:Nokia E7拍摄的201本书的封面图片(多视角拍摄,各6张),共1.3GB; Turin市180个建筑的视频图像,拍摄的camera有Galaxy S、iPhone 3、Canon A410、Canon S5 IS,共2.7GB
  11. Stanford Mobile Visual Search,Stanford Mobile Visual Search Dataset,2011.2,stanford提供,包括8种场景,如CD封面、油画等,每组相关图片都是采自不同相机(手机),所有场景共500张图;以后又发布了一个patch数据集,Compact Descriptors for Visual Search Patches Dataset,校对了相同patch。
  12. UKBench,UKBench database,2006.7,Henrik Stewénius在他CVPR06文章中提供的数据集,图像都为640x480,每个group有4张图,文件接近2GB,提供visual words。

初步版本,水平有限,有错误或者不完善的地方,欢迎大家提建议和补充,会一直保持更新,本文为专知内容组原创内容,未经允许不得转载,如需转载请发送邮件至fangquanyi@gmail.com 或 联系微信专知小助手(Rancho_Fang)

敬请关注http://www.zhuanzhi.ai 和关注专知公众号,获取第一手AI相关知识

成为VIP会员查看完整内容
微信扫码咨询专知VIP会员