Incorrect face boxes in face/recognize endpoint #106
Description
When I try to recognize multiple faces in the image with v1/vision/face/recognize endpoint it returns incorrect face boxes.
To reproduce this bug no faces should be registered in deepstack.
In my opinion, this bug occurs due to misspelling in https://github.com/johnolafenwa/DeepStack/blob/dev/intelligencelayer/shared/face.py#L348