Timeout when registering first face (code 500)

Hey there,

I already posted this in the home assistant forum, but I figured this question is probably better suited here, since I’m just running deepstack in a docker container, without any home assistant components.Hopefully someone here can point me in the right direction

When trying to register a face with deepstack, I run into timeout errors.

The object detection works if I use the VISION-DETECTION=True env with a picture of myself (detects a person). But changing to VISION-FACE=True and then running the python example with the same image, the response just a outputs a “failed to process request before timeout.” after one minute. I can see the request in the docker log after one minute

[GIN] 2021/04/15 - 12:33:20 | 500 | 1m0s | 172.17.0.1 | POST /v1/vision/face/register

I also increased the timeout to 3 minutes, still no luck.

I can successfully query the api for face/list, which of course returns no faces, since, well, the training of the face fails. But that would indicate the api is working in general. I also tried to remove the volume bind without any change. But the faceembedding.db file was created with the volume bind, so it can’t really be a permission issue either.

Any ideas why it’s not registering a new face?

EDIT: I overlooked someone having the same issue on the HA forums. For anyone else looking for this problem, use the deepquestai/deepstack:cpu-x5-beta docker image to solve it.

1 Like