Project Database

Intelligent Inspiration Image Generation Platform for Fashion Design

Generative Adversarial Networks (GAN), which is a recent advance in artificial intelligence (AI)-powered technologies used for generating new (non-existent) objects, can provide the opportunity to realise fashion design through AI. However, the fashion industry has been lagging behind other fields in leveraging AI technology. In this project, we propose to design and develop an intelligent inspiration image generation platform (I3GP) to support fashion product development. The proposed platform uses artificial intelligence (AI)-powered fashion image generation (FIG) technology to support the virtual interactive discussion among the internal and external participants involving in the fashion product development (e.g., fashion merchandisers, fashion product designers, and buyers) via video conferencing or face-to-face meeting. I3GP can generate numerous new fashion design images that users can modify according to specific parameters (e.g., pattern, collar, and season) and predict the design popularity. Users can also automatically record meeting notes with the relevant images during virtual discussion to support fashion product development.

Project No. (ITC) ITP/038/20TP
Project Type Platform
Project Co-ordinator Prof Eric W T NGAI
Research Institute The Hong Kong Polytechnic University
ITF Amount HK$ 4,698,210
Start Date DD/MM/YY 2021-01-25
Completion Date DD/MM/YY 2023-01-24
Sponsors
PD Garments Limited
DIS International Group Limited
Jos Global Limited
Awake Design Company Limited
852 Design Limited
Find 2 Meals Limited
Wavelane Technology Limited
Dragon Computer Electronics Co.