jinlife / synology_photos_face_patch Goto Github PK
View Code? Open in Web Editor NEWSynology Photos Facial Recognition Patch
License: GNU General Public License v3.0
Synology Photos Facial Recognition Patch
License: GNU General Public License v3.0
Hi,
I installed the latest patch but face recognization is still not working.
Is it possible te reverse/remove te patch by any change?
Thanks, Benjamin
J1900 物理机安装 918 7.2 64570 -3 。0489版本photo,按照 https://wp.gxnas.com/11592.html/comment-page-4 操作,重建索引总是会卡住,重启会继续进行索引,面部剩余2张,主题 12000左右,就再也不会索引了,彻底卡死。重启无效,只能重新索引,还会继续卡住
Would the libsynosdk patch also allow remote NFS/CIFS shared folder be used in DSM6's Moments?
DSM 7.0.1-42218 Update 3
DS3615XS
hp gen7
Synology Photos
Before the patch dmesg
print similar to the following output
[28876.461586] traps: synofoto-face-e[2959] trap divide error ip:7f534ae038f2 sp:7f5354a25580 error:0 in libMKLDNNPlugin.so[7f534acc3000+10dd000]
[28882.560670] traps: synofoto-face-e[3068] trap divide error ip:7fda58a828f2 sp:7fda5d8fb580 error:0 in libMKLDNNPlugin.so[7fda58942000+10dd000]
[28888.805445] traps: synofoto-face-e[3097] trap divide error ip:7f109ea438f2 sp:7f10a78d1580 error:0 in libMKLDNNPlugin.so[7f109e903000+10dd000]
[28894.808466] traps: synofoto-face-e[3129] trap divide error ip:7f63b9bff8f2 sp:7f63c2980580 error:0 in libMKLDNNPlugin.so[7f63b9abf000+10dd000]
[28936.600971] traps: synofoto-face-e[3155] trap divide error ip:7fc0ea0d58f2 sp:7fc0f2eb1580 error:0 in libMKLDNNPlugin.so[7fc0e9f95000+10dd000]
[28943.060527] traps: synofoto-face-e[3250] trap divide error ip:7fb77117e8f2 sp:7fb775f5e580 error:0 in libMKLDNNPlugin.so[7fb77103e000+10dd000]
Before the patch cat /var/log/messages
print similar to the following output
2022-04-21T20:35:16-07:00 xxxx coredump[2274]: Process synofoto-face-e[2248](/volume2/@appstore/SynologyPhotos/usr/sbin/synofoto-face-extraction) dumped core on signal [8]. Core file [/volume1/@synofoto-face-e.core.gz]. Cmdline [/var/packages/SynologyPhotos/target/usr/sbin/synofoto-face-extraction ]
2022-04-21T20:35:18-07:00 xxxx coredump[2274]: Core file [/volume1/@synofoto-face-e.core.gz] size [19517704]
2022-04-21T20:35:22-07:00 xxxx coredump[2336]: Process synofoto-face-e[2310](/volume2/@appstore/SynologyPhotos/usr/sbin/synofoto-face-extraction) dumped core on signal [8]. Core file [/volume1/@synofoto-face-e.core.gz]. Cmdline [/var/packages/SynologyPhotos/target/usr/sbin/synofoto-face-extraction ]
2022-04-21T20:35:24-07:00 xxxx coredump[2336]: Core file [/volume1/@synofoto-face-e.core.gz] size [19517825]
After patch, Synology Photos Facial Recognition is working properly
I've deployed DSM 7.2 on a physical machine with Internal Graphics, but the synology photo facial recognition is not working.
I've tried to using this patch to run facial recognition, but failed.
How can I rollback this patch?
DSM 7.1.1-42962
Photos 1.3.3-0330
ESXI 6.7U3:4核 CPU,8G RAM,无核显
替换了so之后,不会卡住”还有N张照片待发现“的进度,但是进度走完之后也没有任何结果出来。全程face-extraction几乎没有CPU的占用开销。
环境:华擎j4105,dsm918-7.1.0-42661,由原来的dsm6.23升级过来的
触发条件:
根据你的说明,替换了so软链接,重启photos,并且重建索引,没有出现人脸识别!
具体错误如下:
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/websocket/socket_notification.cpp:22 send index task failed. io error
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/io/channel.cpp:79 channel[/run/synofoto/js-server-backend.socket] construct failed: connect: No such file or directory
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/websocket/socket_notification.cpp:22 send index task failed. io error
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/io/channel.cpp:79 channel[/run/synofoto/js-server-backend.socket] construct failed: connect: No such file or directory
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/websocket/socket_notification.cpp:22 send index task failed. io error
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/io/channel.cpp:79 channel[/run/synofoto/js-server-backend.socket] construct failed: connect: No such file or directory
2022-11-02T22:29:43+08:00 nas synofoto-task-center[24161]: /source/synofoto/src/lib/websocket/socket_notification.cpp:22 send index task failed. io error
pve虚拟机 j4125cpu 直通核显给黑群晖ds918+
可以用jellyfin做硬解
问下有办法直接调用gpu做识别吗
我是I5 3470 CPU
物理机的时候 人脸识别是完全ok的(因为可以调用核显)。
做成虚拟机后就完全失效(核显没有直通)
换过 libsynophoto-plugin-detection.so 文件也是无效。
请大神看看这个文件是不是也有能禁用GPU的地方呢?这是 Momets的人脸识别
也是6.几的群晖。
使用补丁后失联,重启后需要重装。试了两次。
j1900 物理机下已经补丁成功,现在想换到PVE虚拟机下,不知道能否支持
大佬,群晖发布了新的0263版本,能否更新一下github呢
大佬,你好。根据教程一步一步做了,结果是人脸识别一直待发现(3张照片,40个小时了)。esxi6.7,dsm7.0。8代i5。
大佬好,
我是unraid下的黑裙918,之前是1.2 安装后人脸有效了一段时间,最近新加的图片都没识别人脸了,但是地点啥的还在更新。
查看日志有大量的错误
synofoto-task-center[30253]: /source/synofoto/src/lib/io/channel.cpp:79 channel[/run/synofoto/js-server-backend.socket] construct failed: connect: No such file or directory 2022-10-15T10:40:13+08:00 DSM synofoto-task-center[30253]: /source/synofoto/src/lib/websocket/socket_notification.cpp:22 send index task failed. io error
试过卸载再重装,更到0317并换上新补丁,停止后再启动并重建索引;然后还另外尝试改名成1.0 ,还是老样子.
请教应该怎么办,谢谢!
synofoto.zip
看到网上说要求Synology Photos套件版本是0224。请问是这样吗,支持最新版本吗?
After upgrading to DSM 7.2.1. and patch Photos 1.6.1-0641 all my face mappings got lost and had to reapply manually.
The same happened to 1.6.2-0710. Any hints?
Hi,
I am on DSM version 7.1.1 U2 and have Photos in version 1.3.0.
I copied your patch in the mentioned directory but I receive the following error in /vat/log/messages:
synofoto-bin-team-library-tool[17496]: /source/synofoto/src/lib/io/channel.cpp:79 channel[/run/synofoto/task-center.socket] construct failed: connect: No such file or directory
The Photos app is not creating any entry within the persons folder.
Any ideas?
Hi! There is a new version of Photos, could you release a patch for it? Thanks!
ds918,42962无法正确识别,是不是群晖屏蔽了修复补丁?
用的rr引导DSM 7.2.1, N100的iGPU看起来还不错 [cat /sys/kernel/debug/dri/0/i915_frequency_info]
系统已打开AME的codec. 但是photos没有人物识别, 大佬我可以打这个补丁吗? 因为rr里介绍这个补丁是强制使用cpu来人物识别,但是n100自带核显,有办法不要强制使用cpu,但仍然开启人物识别功能吗?
例如您给的链接里面:
./PatchELFSharp "/usr/lib/libsynosdk.so.7" "SYNOFSIsRemoteFS" "B8 00 00 00 00 C3"
# support face and concept
./PatchELFSharp "/var/packages/SynologyPhotos/target/usr/lib/libsynophoto-plugin-platform.so" "_ZN9synophoto6plugin8platform20IsSupportedIENetworkEv" "B8 00 00 00 00 C3"
# force to support concept
./PatchELFSharp "/var/packages/SynologyPhotos/target/usr/lib/libsynophoto-plugin-platform.so" "_ZN9synophoto6plugin8platform18IsSupportedConceptEv" "B8 01 00 00 00 C3"
# force no Gpu
./PatchELFSharp "/var/packages/SynologyPhotos/target/usr/lib/libsynophoto-plugin-platform.so" "_ZN9synophoto6plugin8platform23IsSupportedIENetworkGpuEv" "B8 00 00 00 00 C3"
注释掉最后一行force no GPU 相关, 能够work吗?
还是说有别的说法,不用这么麻烦直接打libsynophoto-plugin-platform.so
gpu也会被调用? 谢谢大佬!!!
Can I use this plugin on DS214+ ? What other models does it work on?
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.