WDM OS: notices for the on-device meal recognition

Model
  Google AIY Food Classifier, food_V1 (TF Hub id google/aiy/vision/classifier/food_V1), quantised
  TensorFlow Lite build. MobileNet V1, 2,023 dishes plus a background class.
  Licence: Apache License, Version 2.0 (apache-2.0.txt in this folder). The embedded metadata of the file
  states "Apache-2.0".
  Source of the file:
  https://storage.googleapis.com/visionkit/aiy/classifier/seefood_V1/2/quantized_input_uint8/713cfd3d5dc83d2500872dc9e605e8812fd77538-model-with-metadata.tflite
  SHA-256 of the bundled file: 03dd6d9129501f97be00775d9e17b5d9ad13be730149352adb8f4ca953b7a650
  Modifications: none. The file is bundled byte for byte as downloaded.

  Model card, quoted: "Do not use this model to predict allergen or nutrition information."
  The card also says the model assumes its input image contains a well-cropped food dish and that it was
  trained on a dataset skewed toward North American foods. WDM OS uses it only to suggest a dish name that
  the user confirms.

Runtime and code
  LiteRT / TensorFlow Lite (inference engine, pulled in by react-native-fast-tflite): Apache License 2.0.
  react-native-fast-tflite: MIT (react-native-fast-tflite-LICENSE.txt in this folder).
  jpeg-js: BSD-3-Clause; its JPEG decoder is Apache-2.0 and its encoder carries an Adobe BSD-style notice (all in jpeg-js-LICENSE.txt in this folder).

Nutrition data
  U.S. Department of Agriculture, Agricultural Research Service. FoodData Central, 2019. fdc.nal.usda.gov.
  (CC0 1.0 Universal.)
  Contains public sector information licensed under the Open Government Licence v3.0.
  (McCance and Widdowson's Composition of Foods Integrated Dataset, Public Health England, 2021.)
  Contains information from Open Food Facts, which is made available here under the Open Database License
  (ODbL). https://openfoodfacts.org