[Feature] object identification?

[flintstorage812]

I have searched the existing feature requests, both open and closed, to make sure this is not a duplicate request.

• Yes

The feature

I would like to add object search to immich. For instance I would like to search car, truck, dog, cat.

What would even level up immich better is if I can train it. Like label things within a photo and then it starts finding similar things kinda like how it finds people.

Platform

• Server
• Web
• Mobile

[bo0tzz]

You can already search "car, truck, dog, cat" with the existing context search. You can also search "a dog in a car next to a truck full of cats" if you so wish, and if you have a relevant picture it will find it. The CLIP search is much better than any object identification would ever be, so there's no reason for us to implement the latter.

[flintstorage812]

I'm in discord learning, thanks for the reply