Key Features

Obstacle Detection

Real-time detection of obstacles using a fine-tuned YOLOv5s model with 27 custom classes including potholes, scooters, and poles.

Adaptive Voice Alerts

Proximity-based audio feedback to help users safely react to nearby hazards.

Scene Description

On-demand narration of surroundings powered by the Qwen2 Visual Language Model, optimized for spatial orientation and environmental awareness.

Accessibility-First Design

Speech-to-Text, and Text-to-Speech making it assessible to visaully impaired, and Multilingual support for inclusive use.