Evaluation and pricing of reusable fashion products

• Main Authors: Anna Margrét Jónasdóttir 1990-, Karl Arnar Ægisson 1986-, Margrét Vala Björgvinsdóttir 1996-
• Other Authors: Háskólinn í Reykjavík
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: Tölvunarfræði; Gervigreind; Tíska; Smáforrit; Application software; Computer science; Artificial intelligence; Fashion; Iceland; Reykjavik University
• Online Access: http://hdl.handle.net/1946/35694

This report outlines the development of our final project Evaluation and pricing of reusable fashion products in BSc. in computer science at Reykjavik University. The group contains three people, Anna Margrét Jónasdóttir, Margrét Vala Björgvinsdóttir and Karl Arnar Ægisson, and the project was worked in collaboration with K3 Imagine in Iceland. The deliverable is a demonstrable web app based on a microservice architecture that enables a user to quickly determine the type, color and brand of an item using machine learning and enable it for resale along with a report of the machine learning process. The objective of this project is to use machine learning to determine the condition and resale price of used products. The aim is to use this technology to reduce the subjective judgement of employees. The system is a cloud-based app where an employee can log in, take a picture of the product and then get information about the color, type and brand of the clothing, percentage of how well the system did predict the tags and full condition and resale price for the item. If the model returns a false positive, then the user is able to modify the product information, and likewise, the user can add in missing values in the product information if the model returned missing values.