Real-Time Vision Explainer: Object Recognition and Multilingual Voice Narration

Abstract

Understanding visual information is a challenge for many users, especially when images contain multiple objects, people, or complex scenes. Existing solutions often provide limited explanations, depend heavily on proprietary systems, or lack multilingual and voice-based accessibility. This creates a gap for users who require detailed, understandable, and inclusive interpretation of visual content. To address this problem, this project proposes a real-time vision explanation system that analyzes images and provides meaningful explanations. The system allows users to authenticate securely and upload or capture images for processing. It detects objects and recognizes people using computer vision techniques, generates a descriptive summary of the image, translates the description into multiple languages, and produces voice narration for better accessibility. The system also stores user history in a structured text format for future reference. By combining vision analysis, language translation, and speech synthesis using free and open-source tools, the proposed solution enhances accessibility, learning, and interaction with visual data for a wide range of users.