{"repo":"wink-wink-wink555/blind_navigation","free":true,"listed":false,"github":"https://github.com/wink-wink-wink555/blind_navigation","clone":"git clone https://github.com/wink-wink-wink555/blind_navigation.git","description":"An AI-powered solution designed to enhance mobility for visually impaired individuals by combining real-time video analysis with voice-guided navigation. Also features a built-in Multi-Agent AI Assistant — users can speak naturally to complete map navigation, adjust system settings, send messages to family.","language":"HTML","stars":39,"topics":["blind-navigation","location-sharing","ollama","qwen","real-time-video-analysis","voice-synthesis","yolov8","deepseek","mcp"],"license":"MIT","category":"mcp-servers","readme_excerpt":"Blind Navigation (ARIADNE) - Your Way Out of the Labyrinth English 简体中文 📹 Demo Video: V1.0.0, V2.0.0 🏆 HONORS & AWARDS (Click to expand) - Jan 2026 Top 10 , Intel Platform Corporate AI Solution Innovation Practice Competition (Global Finals) - Dec 2025 Top 20 , Intel Platform Corporate AI Solution Innovation Practice Competition (Preliminary Round) — Advanced to Global Finals - Aug 2025 Third Prize (National Level) , Chinese Collegiate Computing Competition (CCCC) - May 2025 First Prize (Provincial Level) , Shanghai Computer Application Competence Competition for College Students --- 🌟 Introduction Travel Assistance System for the Visually Impaired (ARIADNE) is an AI-powered navigation system designed for visually impaired individuals. It combines computer vision and AI to identify tactile paving through real-time video analysis and provides intelligent voice guidance. The system also features a built-in Multi-Agent AI Assistant — users can speak naturally to complete map navigation, adjust system settings, send messages to family members, and more. Additional features include real-time location sharing to enhance travel safety and independence. Tech Stack - Frontend : HTML5, CSS3, Vanilla JavaScript - Backend : Flask (Python 3.8+) - AI Models : - YOLO (You Only Look Once) - Tactile paving detection - AI Text Model (flexible): supports both cloud APIs (DeepSeek, OpenAI, DashScope/Qwen, or any OpenAI-compatible endpoint) and local Ollama models — configurable per user in th","default_branch":null,"files":null,"tree":[],"storefront":"/r/wink-wink-wink555","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/wink-wink-wink555/blind_navigation/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}