Split Object/Face Recognition

[tknobi]

Hi,

For me, recognizing objects and recognizing faces are two completely different things. While objects are simply recognized, faces have to be detected, a person has to be recognized, suggested, labeled, filtered, clustered, ....

So should this be split?

Cheers
tknobi

[meichthys]

Yes, i agree. I would think that any application that has face detection will also have object detection, but the quality of object detection could still vary widely.

We should probably also split out some other 'Features' like Mobile App and Desktop App. I'll create separate issues for those.

[meichthys]

• If we split Object/Face recognition, we will need to update features.md

[rmspot]

I think (not an expert tough) that object vs face detection is something like common noun vs proper noun: face detection tells you this is Paul, Mary, etc. while object detections tells you this is a face, a head, a car, a tree ... then object detection could be more specialized, instead of telling this is a car it would tell this a Ford, a Jaguar, ... and the exact model if it can