lolishinshi / imsearch Goto Github PK
View Code? Open in Web Editor NEW基于特征点匹配的相似图片搜索
License: GNU General Public License v3.0
基于特征点匹配的相似图片搜索
License: GNU General Public License v3.0
最新版
HZ-E5-256G-10并行-230G(HDD)
[2021-12-01T15:47:09Z DEBUG imsearch::index] knn search time: 3.67s
[2021-12-01T15:47:09Z DEBUG imsearch::index] knn search time: 4.17s
[2021-12-01T15:47:09Z DEBUG imsearch::index] knn search time: 4.20s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.24s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.26s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.26s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.27s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.28s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.29s
[2021-12-01T15:47:10Z DEBUG imsearch::index] knn search time: 4.29s
[2021-12-01T15:48:54Z DEBUG imsearch::imdb] search time: 108.33s
[2021-12-01T15:48:56Z DEBUG imsearch::imdb] search time: 111.20s
[2021-12-01T15:48:57Z DEBUG imsearch::imdb] search time: 111.74s
[2021-12-01T15:48:58Z DEBUG imsearch::imdb] search time: 112.45s
[2021-12-01T15:48:58Z DEBUG imsearch::imdb] search time: 112.55s
[2021-12-01T15:48:59Z DEBUG imsearch::imdb] search time: 113.93s
[2021-12-01T15:49:00Z DEBUG imsearch::imdb] search time: 114.26s
[2021-12-01T15:49:00Z DEBUG imsearch::imdb] search time: 114.42s
[2021-12-01T15:49:00Z DEBUG imsearch::imdb] search time: 114.89s
[2021-12-01T15:49:01Z DEBUG imsearch::imdb] search time: 115.48s
在E5-1650V3 256G HDD机子里
使用 start-server http模式加载 230G index 和 22G 数据库(database)(数据库已 clear-cache 清除特征点)
再使用10个并行测试搜索,log如上
index已经缓存在内存里,knn搜索时间没什么问题
但总时间却过于长,应该是「从数据库中查询特征点对应的本子」,并且数据库存放在HDD读取过慢的问题
1.把数据库(database)存放在nvme盘
但租用的服务器没有
2.使用内存创建虚拟硬盘,把数据库(database)放进去
目前临时使用该方法解决中,但由于是使用tmpfs创建内存虚拟硬盘效率不佳,并且只能手动
3.我建议新增个可选项,把数据库(database)也缓存在内存里
最新版
添加48w图片时的一半出现
[OK] Add /www/nhentai_jp/nh2/[315879][brother pierrot]/091.png
[OK] Add /www/nhentai_jp/nh2/[315879][brother pierrot]/146.png
[OK] Add /www/nhentai_jp/nh2/[315879][brother pierrot]/001.jpg
Segmentation fault
新建一个数据库,添加这个 /www/nhentai_jp/nh2/[315879][brother pierrot]/
文件夹是没问题
新建一个数据库,添加这个 /www/nhentai_jp/
整个文件夹,也在中途出现 Segmentation fault
错误中断
但中断文件的文件夹不是 /www/nhentai_jp/nh2/[315879][brother pierrot]/
,而是其他
退回上上旧版 https://github.com/lolishinshi/imsearch/tree/06b4c3705ec5554c7429923dbc5de5bb6fb85645
再新建一个数据库,添加这个 /www/nhentai_jp/
整个文件夹,未出现中断错误问题
第一次的imsearch add-images train和后面的imsearch add-images DIR是什么关系?看起来train的图片也会被搜索,区别是?
执行多次imsearch add-images train添加训练数据,同名图片会覆盖掉吗?imsearch add-images DIR会被用于训练吗?
fatal error: faiss/index_factory.h: No such file or directory
cargo:warning= #include <faiss/index_factory.h>
只能从源码安装吗?我用conda安装的rust和faiss
conda create -n imsearch -c conda-forge rust python=3.7
conda install -c pytorch opencv faiss-gpu cudatoolkit=10.2
出现了上面那个报错
比如QQ群什么的
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.