
IsloPol
Circular Economy2025Turning EPS waste into local circular value.
Back to projects

FOCUS, Foot-based Control for User Support, is a project to develop a ready-to-use AI-based foot gesture recognition system using only a standard RGB camera. The solution enables intuitive, hands-free interaction with digital systems through simple foot movements, without wearables, specialized hardware, or any type of physical contact.
Designed for flexibility, FOCUS can be deployed in a wide range of environments where traditional interaction methods are limited, inconvenient, or simply not the best option. The system works with most standard RGB cameras available on the market, enabling easy integration into existing setups without the need for specialized hardware.

Human–machine interaction has long relied on hands to operate keyboards, touchscreens, buttons, and controllers. But in many real-world situations, hands are busy, unavailable, or interaction needs to be faster, simpler, or more natural.
FOCUS introduces a new interaction modality that complements existing ones. Foot gestures can be used as an alternative when hands cannot be used, as a complement to enhance speed and efficiency, or as a fallback in environments where voice or touch is not viable.
This creates a more adaptable, inclusive, and resilient way to interact with digital devices across accessibility, healthcare, industrial, and immersive application contexts.





One of the most impactful applications of FOCUS is in the health and accessibility domain. For individuals with reduced mobility in the arms or hands, interacting with digital systems can be difficult, slow, or even impossible. FOCUS provides a new simple, reliable, and intuitive channel of interaction, enabling users to perform tasks using foot gestures.
This is especially valuable in assistive technologies for people with upper limb impairments, rehabilitation and therapy environments, accessible interfaces for public or private digital systems, and everyday interactions where simplicity and independence matter.
While gesture recognition has advanced significantly, most existing solutions focus on hands or full-body tracking. These approaches do not translate well to scenarios where only the lower body is visible, the camera perspective is floor-level or top-down, users wear different types of footwear, and lighting and environmental conditions are highly variable.
Additionally, there are no widely available datasets or pre-trained models specifically designed for foot gesture recognition in real-world conditions, making FOCUS both a technical and practical step toward a missing interaction technology.
FOCUS uses a modular AI pipeline running on edge devices to recognize foot gestures in real time. The system combines the detection of feet or footwear in the camera stream, extraction of foot keypoints and movement features, and analysis of temporal patterns to identify gestures.
This allows the system to interpret actions such as taps, swipes, or directional movements and map them to commands in a digital interface. Unlike sensor-based or radar-based approaches, FOCUS is designed for practical deployment at scale, using widely available RGB cameras and a flexible API-based integration strategy.





FOCUS introduces a vision-based, non-contact interaction system that is camera-only, non-intrusive, flexible, real-time, and scalable. It is designed to integrate with existing systems without requiring specialized devices, body-mounted sensors, or changes to the physical environment.
Uses standard camera input only
Wearables or body-attached sensors required
Ready for integration into digital systems
Real-time gesture interpretation on edge devices

ENFIELD Project Context

