Our research
/ 01
Integration of Computer Vision in Fashion
The CAFI Research Group is dedicated to unlock unlimited potential and possibilities through integrating Computer Vision in the fashion industry.
We focus on the development of innovative systems and deep learning models for computer vision tasks like classification, object detection, segmentation, human/clothing paring, pose estimation, image & item recognition and retrieval. These technologies enable computers to recognize, process, analyse and understand digital images or other visual inputs, then derive meaningful information from them. These intelligent image understanding and analysis techniques can be used for novel fashion applications like body shape modelling and fashion recommendations, helping fashion professionals improve customer’s digital experience, derive e-commerce retailing strategies, conduct consumer behavioural & fashion trend studies and beyond.
research paper
research projects
/ 02
computer aided design (CAD)
The CAFI Research Group has years of in-depth research experience in computer aided design (CAD) technologies and this has always been at the heart of our research.
Our group explore how advanced computer technology can be adopted across the fashion value chain. We develop transformative algorithms, systems and platforms to revolutionize how textile/fashion products & experiences are designed and produced. Our researches range from smart fashion graphic & design synthesis and generation, intelligent pattern planning, pattern generation and auto-grading, to the use of blockchain technology for product mass customization and more. We aim to optimize the value chain by introducing innovative perspectives in how fashion is designed and produced.
research projects
/ 03
smart fitting & sizing
The CAFI Research Group is an award-winning research team in the area of intelligent fitting and sizing.
Our technology can digitally reconstruct the shape and size of a person with 2 full body pictures within 10 seconds, hence creating a personalized avatar. Other than body shape modelling, CAFI research group also works on face and foot reconstruction for a truly personalised avatar ready for the metaverse. With advanced big data analytics, highly accurate body measurements are generated for smart and personalized recommendation on size and shape. This opens new opportunities to better online shopping experience and new facilitating model for bespoke services.
We want to push the limits for faster, more convenient and accurate intelligent fitting and sizing for the fashion industry – whether it is to find the best fit ready –made products, or to offer made-to-measure service in a mass scale without geographical limitations.
research projects
Project period: 01/08/2020 – 31/08/2023
/ 04
mixed reality
At the CAFI Research Group, we explore and introduce a new level of interaction between fashion products and customers .
We focus on blending both Augmented Reality (AR) and Virtual Reality (VR) technologies to enable more engaging, interactive and immersive product viewing and brand storytelling. Creation of a personalized avatar will just be a few clicks away, which can be customised in look, expression and motions Customers can virtually view the products and their intricate details in full 360 degrees. They can also try on and interact with these hyper-realistic digital twins in real time via their phones or digital device.
We believe the future of fashion is about this exciting multisensory experience which we can interact with data and digital elements in the real world.
research projects
Project period: 1/1/2009 – 30/06/2011
/ 05
Sustainability and Integration (SI)
The CAFI Research Group also has a significant impact on sustainability.
We focus on sustainability and integrationThrough optimizing the product development and production practices with computer-aided innovative technologies, efficiency is enhanced and redundant processes are reduced, leading to smarter use of materials and energy, hence conserving resources and reducing carbon footprint. For example, technology developed like smart fitting and sizing tools promotes more accurate product-user fit, they not only give much convenience to the users but also greatly reduce the chance of wastage caused by inappropriate sizing. The use of mixed reality technologies elevated designers and consumer experience with hyper-realistic simulation, as a result can help reduce the reliance on actual samples and hence reduce wastage on sampling. With tools like the magic wardrobe, consumers can virtually better organize and make use of the fashion products they already own conveniently.
We are dedicated to promoting a sustainable consumption culture in which one can make the best use of resources and reduce the tendency of impulsive purchases.
research projects
Project period: 1/1/2009 – 30/06/2011