mrousavy / vision-camera-image-labeler

VisionCamera Frame Processor Plugin to label images using MLKit Vision

Home Page:https://github.com/mrousavy/react-native-vision-camera

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Why the ML show only a few labels?

patras1 opened this issue · comments

commented

I have installed the labeler but it shows only a few labels no matter what the camera sees: Metal, curtain ,pattern, food.
Does anybody know how can I make it show labels for food item?
Let say banana, apple,egg etc...

Thanks,
Dan

Hey I think this is due to the fact that the input format (RGB/YUV) or the orientation (landscape/portrait) is wrong.

I didn't update this repo in a while and I think the future should be the TensorFlow Lite package I'm drafting up for V3 launch.