AI-driven image analysis

La Trobe researchers are working to understand and interpret complex visual data through artificial intelligence and computer vision

In an era where digital images reach trillions per year globally, the need to efficiently process and analyse visual data is growing in importance.

This is why Dr Zhe Chen is dedicating his research to addressing the current challenges of understanding and interpreting complex visual data through artificial intelligence and computer vision.

“Artificial intelligence is the simulation of human intelligence in machines that are programmed to mimic human-like cognitive functions such as learning, problem-solving, decision-making and perception,” explains Dr Chen.

“Computer vision is a subfield of artificial intelligence, enabling computers to interpret and understand visual information from the real world such as image, object and facial recognition.”

“It is particularly crucial in areas such as autonomous driving and smart robotics, but its impact can extend into other fields like healthcare and public safety.”

The market for artificial intelligence-driven image analysis is anticipated to grow into a multi-billion-dollar sector by 2025, further highlighting the demand for this technology. Despite this, there is a pressing need for computer vision models that are both robust and cost-effective.

"Balancing the robustness of these models with affordability poses a significant challenge, especially in cases where real-time processing and decision-making is required," says Dr Chen.

“Existing computer vision models also struggle to adapt to changing environments which requires extensive model re-training and recalibration, significantly limiting their versatility and scalability.”

"To address this issue, our research has been focused on integrating contextual information and simplifying the interpretation of otherwise hard-to-recognise objects so that models can easily adapt to environmental changes.”

“This has significantly reduced the need for model re-training, thereby increasing the cost-effectiveness and practical application of these models.”

Dr Chen now hopes to take these findings and develop practical, real-world computer vision models for a range of sectors.

“One such focus will be the healthcare sector where, through making sophisticated computer vision models that are accessible, practical and cost-effective, I hope to help improve diagnostic processes and patient care.”