johnolafenwa
Docker CPU face recognize request with random image always produces 500
If the Docker version is used with CPU, any request to recognize face with a random image and no known faces almost always results in error 500 with a 1m timeout.
This isn't only me, you can find it widely reported:
-
https://github.com/robmarkcole/HASS-Deepstack-face/issues/39
-
https://forum.deepstack.cc/t/vision-face-times-out-on-latest-july-2021/1038
-
https://www.reddit.com/r/BlueIris/comments/obmwy8/deepstack_face_troubleshooting_docker/
I found the only solution is to use the deepquestai/deepstack:cpu-x5-beta image. No others work. I found that solution from here -- https://forum.deepstack.cc/t/timeout-when-registering-first-face-code-500/871
Please fix this.
This is similar to https://github.com/johnolafenwa/DeepStack/issues/92. I am having the same issue on Mac Mini M1 using the latest (and also tried the first) arm image.
I have the same issue. Only thing to resolve issue is to restart container. Then it works for me for a brief moment, then back to 500 errors and 1 m timeouts. I am using Deepstack 2022.01.01...
Hello @parhedberg @JurajNyiri @automaton82 , sorry to hear you are having this issue. We would fix this and release a an update in the next couple of days. Thanks for your patience
+1 .. Face Detection on Docker (GPU & CPU Versions) doesn't work and errors out with a 500.. Same behavior for all 2021 and the current release.
I have the same issue on the Jetson Nano, I hope some how this gets taken care of.
I'm getting a lot of 500 errors on windows native CPU for face detection.
not only on docket not only on face recognition, CPU does that straight MSI install on windows with Object recognition.. not even answering with err 500 back to client with ERR explaining problem,.. just no answer :\
I got 500 only on my jetson AGX but when i run on my desktop i got 200 , however it seem not detect anything just result 200 and give blank answer .. but jetson AGX sure 500 no matter what ..
