Challenge Project

Snaq on Eschernode

Estimate food portion size, nutritional value, contents automatically

Pitch deck

Use Cases

  • People will know the nutritional composition instantly.
  • Identify the dish, especially helpful when people are travelling, so that they can know what they are eating. eg. Simply take a picture at a buffet and know what it is.
  • Helpful to people following a diet.

First steps.

  • The data set we were provided is a 700 pictures each of three different classes. Identify these classes automatically.
  • Available Classes:
  • Bowl plate
  • Regular plate
  • Soup plate

Training using Eschernode

  • Trained a deep learning classifier on Resnet18 architecture.
  • Trained on a 60GB machine for 5 hours.

Results

  • Best Cross entropy error of about 0.863 after training for 42 epochs

Edited

30.01.2018 08:22 ~ oleg

Challenge

Event finished

Event started

 
Contributed 6 years ago by oleg for Open Food Hackdays Lausanne

Connect to our community on Mattermost | Twitter | Facebook


Creative Commons LicenceThe contents of this website, unless otherwise stated, are licensed under a Creative Commons Attribution 4.0 International License.