Food Waste Analytics and Visual Feedback for Benchmarking and Behavior Changes
Project Description
We aim to reduce the food waste generated in HKUST by first understanding what sort of food are left unfinished through automated video analysis. Then, the information learnt is utilized to spread awareness for promoting behavior changes. This project is a part of the HKUST Smart Sustainable Campus (SSC).
Supervisor
CHENG, Kwang Ting
Co-Supervisor
QU, Huamin
Quota
3
Course type
UROP1100
Applicant's Roles
1. Collaborate with PhD and FYP students in developing a living lab in HKUST food premises.
2. Develop, implement, and improve the automated video-processing tool, e.g., video-based food semantic segmentation, to correctly classify and estimate the volume of the leftovers.
3. Visualize of the information obtained from the processing the video data for two stakeholders, the public and caterers.
4. Propose and implement tools or activities, e.g., campaign, to promote food waste awareness and behavior changes. For instance, a community-based mobile app to encourage people to finish their food.
2. Develop, implement, and improve the automated video-processing tool, e.g., video-based food semantic segmentation, to correctly classify and estimate the volume of the leftovers.
3. Visualize of the information obtained from the processing the video data for two stakeholders, the public and caterers.
4. Propose and implement tools or activities, e.g., campaign, to promote food waste awareness and behavior changes. For instance, a community-based mobile app to encourage people to finish their food.
Applicant's Learning Objectives
1. Obtain first-hand experience in designing and developing an AI-based end-to-end solution to addressing a real-life and meaningful application for a societal problem
2. Learn how to design products or practical solutions based on the different stakeholders’ need and want.
2. Learn how to design products or practical solutions based on the different stakeholders’ need and want.
Complexity of the project
Moderate