Android App that uses the CameraX Image Analysis Use-Case and a ML model to classify food in the images captured by the phone back camera.
The TensorFlow Lite food classification Model by Google used is based on MobileNet V1 and was trained to recognize food dishes from images on a dataset skewed toward North American foods. The training set includes entrees, side dishes, desserts, snacks, etc. For details of the model architecture, see MobileNet V1.
Screenshots of the app:
