Advancing Face Recognition Technology: Exploring 2D and 3D Synthetic Face Image Generation using AI

Applied Recognition Corp. (ARC) a Canadian software company, which develops face recognition technology solutions, which are engineered to be both accurate and practical, allowing your organization to upgrade to biometrics simply and inexpensively. ARC relies on high-quality training data sets of human face images to train their machine learning models and ensure the accuracy of their image detection algorithms. However, they face a significant challenge in terms of the availability of high-quality data for training and testing purposes. To tackle this challenge, ARC is experimenting with state-of-the-art artificial intelligence (AI) image generation tools. ARC is collaborating with WIMTACH CEDA to research and analyze the limitations and challenges associated with generating synthetic face images using various AI image generation tools.

ARC is continuing its collaboration with WIMTACH to address the critical challenge of high-quality training data for its machine learning models. To ensure accuracy and practicality, ARC relies on robust datasets of human face images, but the availability of such data remains a key obstacle.

Building on previous research conducted by ARC, this project focuses on advancing the use of AI image generation tools to create synthetic face images. The project will delve deeper into the limitations, challenges, and opportunities of synthetic data generation, aiming to optimize the performance of ARC’s facial recognition algorithms.

A key outcome will be a comprehensive design document outlining the control features and parameters—such as image diversity, resolution, realism, and bias mitigation—that influence the quality and utility of synthetic images. This project aims to provide actionable insights to enhance the effectiveness of ARC’s biometric solutions while ensuring cost-effective, scalable, and ethical data practices.

Faculty Supervisor:

Tenzin Jinpa;Sreekumar Radhakrishna Pillai

Student:

Partner:

Applied Recognition Corp

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

Centennial College of Applied Arts and Technology

Program: