David Jakowenko

Hey @bigbangus, thanks for letting me know about this. I've responded on the CompreFace issue and will keep an eye on it if there's something I need to update.

Thank you for the suggestion and checking out Double Take. In the meantime, you could use a [Chrome extension](https://chrome.google.com/webstore/detail/image-downloader/cnpniohnfphhjihaiiggeabnkjhpaldj) and save all the images locally, then upload them in bulk...

> somebody is using double-take with the heaviest model. great. but it is eating memory... or is it double-take? -> [exadel-inc/CompreFace#569](https://github.com/exadel-inc/CompreFace/issues/569) Double Take shouldn't be. It's a pretty light weight...

The `double-take/cameras/` topic does include the unknown and array of misses. Would this work? 

You can feed the API manual images to observe what the results are. Here's a movie poster image that produced "misses" for me when I used it with my data...

What are your Frigate camera names called? Does the correct camera name show up on the Match tab on the UI? What detector are you using?

Can you paste the value of the entities attributes? Also, what is the state value of it? I've noticed DeepStack returns 0% a lot on misses, you may want a...

Oh. I see what you are doing. You're using the camera entity to trigger the notification, and the state of that shows how many people were seen. The example in...

> I wasn't sure if my response might've been confusing, but I do still need help getting notifications setup of specific people. Once I know how to do it for...

Thanks for the additional info @bigbangus, I do look for `after` if it exists, but I'm guessing the initial match came during the `before` event, which at that time had...