Reproduce “SANIP: Shopping Assistant and Navigation for the visually impaired” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shubham Deshmukh, Favin Fernandes, Amey Chavan, Monali Ahire, Devashri Borse, Jyoti Madake
SANIP, developed by researchers at Vishwakarma Institute of Technology, presents an assistive system for the visually impaired that integrates custom object detection, text recognition, and barcode/QR code decoding with offline text-to-speech feedback. The prototype successfully identifies common grocery items, reads labels, and decodes product information, enhancing independent shopping capabilities.
Key insights
- Integrating multiple computer vision tasks (object, text, and barcode detection) within a single system offers a more comprehensive and practical assistive tool for visually impaired shoppers.
- The selection of YOLOv4-tiny for object detection and Pyttsx3 for offline text-to-speech prioritizes real-time performance, ease of implementation, and reliability independent of internet connectivity.
- Creating a tailored dataset for specific, frequently encountered shopping items enables the system to focus on immediate practical utility for users.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.