That seems indeed to be the issue, changed into pushover message without image and that works fine. Anyway, thanks for your help and great apps. A well deserved donation for the good work is on the way.
1 Like
I made a new version 1.2.0. This has a major new feature: You now get the detected face as a cropped image in the trigger flow card.
- Added cropped face image tag in trigger card.
- Small fixes.
Available now as test version: https://homey.app/a/com.gruijter.facepp/test/
1 Like
I just installed the normal version and looks fun yo use!
Thanks for all your work.
Im going to test the test version!
1 Like
Tested a few times and it makes the app on pauze.
It removes the token and secret and I have to make a new one.
Maybe it has to do with to much snapshots?
It makes more shots then the normal version
Hi @Gruijter,
You have a personal message with the log file.
It seems that after a while it recovered itself automatically in the 1.1.7.
Version 1.2.1 was released as stable to the app store.
It adds the captured face image to the trigger flowcard and resolves a memory leak.
Have phun 
Love the new feature but as always that leads on to wanting more 
Would it be possible to add a device that can show the last detected face?
Maybe even one device per source (not sure how you would do that)?
Or maybe one image per source such that the single device registers an image for each source (by that I guess I mean camera)?
I do something similar in the ONVIF app where there is an image that is captured when motion is detected plus a āliveā image.
Always wanting moreā¦ I know the feeling.
As you might have noticed the face++ app does not use devices, and I want to stay away from that for this app. Secondly I found out that I need to ādestroyā the face image streams at some point in time to avoid an indefinite memory usage (it is the memory leak that I fixed in 1.2.1). I destroy the stream now after 3 minutes.
I thought of creating a global image token that always contains the last detected face, but since the stream is automatically destroyed, that image would only be available for 3 minutes.
So you see, not so easyš
Maybe in the future I will have a brain wave and can come up with a solution. Would a global image token fit your needs?
A global image would be fine for me at the moment as I only use it on one camera. But that might change in the future
Maybe it could be another tab in the config page for the last detected face?
What you have given us is fantastic so Iām not complaining about this missing feature.
What is your use case for having the last detected face available in Homey? I send all snapshots to telegram, so there I actually have all snapshots that were taken for the last few days, including timestamps and source id.
Iāve not used telegram so will look into that.
It is just so I can have somewhere to see the last detected face as at the moment once I dismiss the notification from my phone the face has gone.
Telegram is your answer then
Error in latest version;
logs
LOG ERROR NO_DETECT
Error: not_found
That is not enough information for me to understand what your issue is. Can you make an issue on github with more background. E.g. In what context does this happen. Thx!
No more info, app stops often and a need of app restart. All data (Api) in app gone before restart.
Memory issue maybe? How loaded is your Homey with apps (memory and cpu load)?
Hey, this seems interesting!
I tried setting it up today to manage my presence and open my lock for me. I get no match and 0 confidence in the log, but the flow still triggers and opens my lock. What am I doing wrong?
2020-08-15 18:56:54 [log] { origin: 'Front Door ', label: āNOMATCHā, confidence: 0, token: āe770f92336cd0e1304951c40bebd0c3bā, quality: 0.006, gender: āMaleā, age: 35, emotion: āsurpriseā, normalglass: false, sunglass: false, mask: false, face_image_token:
In your flow you need to add a condition. E.g. AND label = yourname AND confidence larger then 80
Did you add your face already through app settings? That is the place where you label your face with a name.
Yes I have added my tokens and faces in app settings.
I will try to assign conditions in And section. I was trying to recreate some flows that were posted above.
Also, how do I make it search for the face multiple times in the flow? For example someone posted an example and they have it search for certain amount of times, but then flow also triggers multiple times. I can have it search NOW image or Motion Event image, also not sure which one to pick.
Edit: I couldnāt find where to set conditions 
do I have to use logic for it?
