Media release

From: The Royal Society

The gait detective – Whether a shuffle, saunter, stride or stomp, AI can identify individuals based on their distinctive walk. A database of over 700 people from different countries was used to train AI models to identify individuals using the unique forces applied to the ground whilst walking. Models trained on individuals in their own shoes had accuracy upwards of 89%, highlighting how gait analysis could play an important role in personalised healthcare and security.

Modelling individual variation in human walking gait across populations and walking conditions via gait recognition

Journal of the Royal Society Interface

Researchers have discovered how our distinctive ways of walking, or gait, can reveal important details about our identity to enhance healthcare and security systems. Using a database of over 700 people from diverse countries, Artificial Intelligence (AI) models were trained to use forces applied to the ground during walking as a personal movement signature. The models were strikingly sensitive to different shoes, walking speeds, and biological traits, shedding new light on how these factors coalesce to define our walking styles. Models exposed to diversity during learning were highly accurate at identifying people in various conditions, highlighting the powerful combination of force signatures and AI for personalised healthcare and security.