The benefits of taking part in adventurous activities are many; particularly, for people with visual impairments. Sports such as rock climbing can improve feelings of skillfulness, autonomy, and confidence for people with low or no vision as they strive to overcome environmental and personal challenges. In this late-breaking work we present Climb-o-Vision, a novel sensory substitution software that utilizes YOLOv5 computer vision object-detection architecture, to aid navigation for rock climbers with visual impairments. Climb-o-Vision uses commercially available and cost-effective hardware to detect, track, and convert climbing hold spatial locations on to the surface of the tongue, via an electrotactile tongue interface. Preliminary testing of the device highlights the possibility of using sensory substitution as a sporting aid for people with visual impairments. Furthermore, it demonstrates the potential for adapting and improving current sensory substitution systems by employing computer vision techniques to filter useful task-specific information to users with visual impairments.
hironobutakagi
commented
2 years ago
Links
Abstract
The benefits of taking part in adventurous activities are many; particularly, for people with visual impairments. Sports such as rock climbing can improve feelings of skillfulness, autonomy, and confidence for people with low or no vision as they strive to overcome environmental and personal challenges. In this late-breaking work we present Climb-o-Vision, a novel sensory substitution software that utilizes YOLOv5 computer vision object-detection architecture, to aid navigation for rock climbers with visual impairments. Climb-o-Vision uses commercially available and cost-effective hardware to detect, track, and convert climbing hold spatial locations on to the surface of the tongue, via an electrotactile tongue interface. Preliminary testing of the device highlights the possibility of using sensory substitution as a sporting aid for people with visual impairments. Furthermore, it demonstrates the potential for adapting and improving current sensory substitution systems by employing computer vision techniques to filter useful task-specific information to users with visual impairments.