iWalkSafe

Basic Information

• Total Award Grant: US$1,575
• Team Size: 3
• Duration: 6-Month
• Role: Lead Developer

Introduction Video (Please turn on the CC for English subtitles)

Problem Addressed

iWalkSafe addresses the significant challenges faced by visually impaired individuals in navigation and mobility. The project focuses on developing a computer vision-based wearable navigation assistance system that leverages miniaturized AI technology. This innovation aims to enhance the safety, independence, and quality of life for visually impaired people by providing a high-mobility solution to mitigate common dangers in their daily environments.

Personal Contributions

As the lead of a development team of three, I contributed 35% of the development work for iWalkSafe, and 100% of the paper writing. My role involved spearheading the implementation of a novel, practical design for a visual assistance system using a wearable RGB camera and an edge AI device. This involved integrating our patented CNN miniaturization method into a critical application like offline navigation assistance, ensuring the system’s high mobility, low power consumption, and maintaining a negligible precision trade-off.

Solutions Developed & Impact

The iWalkSafe system is a breakthrough in wearable navigation assistance, utilizing a patented miniaturization technology to deploy AI models on edge devices. This design enables advanced computer vision tasks such as object detection, people tracking, and facial recognition while operating offline in real-time with low power consumption. The system informs users through audio output and haptic feedback for danger avoidance or environment perception, significantly enhancing the safety and independence of visually impaired individuals. The success of this project demonstrates the practical application of miniaturized edge AI devices in personal assistive equipment.

Learning Outcomes & Reflections

Leading the iWalkSafe project was a remarkable experience that underscored the potential of edge AI and miniaturization technologies in addressing real-world challenges. It honed my skills in computer vision, AI development, and team leadership. This project highlighted the importance of interdisciplinary collaboration and user-centric design in developing assistive technologies. It was particularly rewarding to see the tangible impact of our work on improving the lives of visually impaired individuals, further motivating me to pursue innovations that bridge the gap between advanced technology and everyday practical applications.

Recognition

Silver Award 
Intelligent SoC Innovative Project Contest
Ministry of Education, Taiwan
Aug 2022