下面要看的文件列表
http://blog.csdn.net/hjimce/article/details/50098483
paper主要創(chuàng)新點:在創(chuàng)新點方面,我覺得主要是把Siamese 網(wǎng)絡的雙分支,合在一起,從而提高了精度
主要講解2015年CVPR的一篇關(guān)于圖像相似度計算的文章:《Learning to Compare Image Patches via Convolutional Neural Networks》,本篇文章對經(jīng)典的算法Siamese Networks 做了改進.
person re-identification dataset :http://robustsystems.coe.neu.edu/sites/robustsystems.coe.neu.edu/files/systems/projectpages/reiddataset.html
http://www.micc.unifi.it/lisanti/source-code/re-id/
Person Re-Identification by Iterative Re-Weighted Sparse Ranking
http://blog.csdn.net/lv26230418/article/details/51312348?locationNum=15&fps=1
An Enhanced Deep Feature Representation for Person Re-identification
http://blog.csdn.net/zdh2010xyz/article/details/53646804
Person Re-Identification Meets Image Search
(market-1501)
A Multi-task Deep Network for Person Re-identification
行人檢索”Deep Ranking for Person Re-identification via Joint Representation Learning”
把圖像拼在一起進行輸入,比如一張圖片是256*128,拼在一起是256*256
Deep Attributes Driven Person Re-identification
http://www.th7.cn/db/mysql/201411/78883.shtml
An enhanced deep feature representation for person re-identification
Constrained Deep Metric Learning for Person Re-identification
總結(jié)
- 上一篇: Shell中的${}、##和%%使用范例
- 下一篇: matlab下删除文件或文件夹