hackquest logo

BrightWalk

Bright Walk is a voice-operated app that assists visually impaired individuals and tourists by providing scene recognition, navigation, multilingual translation, and itinerary suggestions in real-time

Videos

Description

Bright Walk is an inclusive Android app designed to empower visually impaired individuals and tourists navigating unfamiliar environments. By integrating multimodal AI (Qwen), OCR, and real-time translation, it addresses critical challenges through a 100% voice-operated interface and support for 13 Indian languages, ensuring accessibility across diverse linguistic demographics.

Key Features

For Visually Impaired Users:

Scene Exploration: Capture images via voice commands and ask iterative questions (e.g., “Describe the left side” or “Read the sign ahead”) for granular insights.

Multilingual Support: Translates text from signs, menus, or labels into 13 Indian languages (e.g., Hindi, Tamil, Bengali).

Dynamic Navigation: Provides audio-guided directions and obstacle alerts.

For Tourists:

Language Barrier Solutions: Translates text from photos into the user’s native language, including regional Indian languages.

Journey Planning: Suggests nearby hotspots (historical sites, restaurants) with cultural context.

Unified Functionality:

Voice-Driven Workflow: Fully operated via natural language commands.

Contextual AI: Retains image context for follow-up queries (e.g., “Find vegetarian restaurants nearby” after scanning a street).

Market Differentiation

Replaces Multiple Apps: Combines OCR, translation, and navigation into one platform.

Localized for India: Supports 13 Indian languages, bridging gaps in rural and urban accessibility.

Inclusive Design: Tailored for voice interaction, eliminating reliance on visual UI.

Impact

Bright Walk bridges independence gaps for 4.95M visually impaired Indians while streamlining travel for tourists. By enabling iterative Q&A, AI-curated journey suggestions, and multilingual support, it reduces cognitive load in foreign environments.

Technologies Used:

Multimodal AI (Qwen) for scene understanding.

OCR (Tesseract) with Indian language support.

Speech-to-text (Google Speech) for voice commands.

This app has the potential to serve 35M+ global tourists annually, fostering inclusivity and seamless exploration.

Progress During Hackathon

It began with a simple idea: “What if one app could guide both a blind person through a bustling Indian market and a tourist lost in Tokyo?” We started by tackling scene recognition—training a multimodal AI (Qwen) to describe environments in real-time. Next, we wove in voice commands, letting users ask, “What’s ahead?” and hear instant replies. But language barriers lingered, so we added OCR and translation, enabling the app to read a Japanese menu and recite it in Hindi. For tourists, we mapped nearby hotspots—temples, cafes, museums—and let them ask follow-ups: “Is there vegetarian food here?” Then came the magic touch: multilingual support for 13 Indian languages, ensuring a grandmother in rural Gujarat could navigate as effortlessly as a tech-savvy traveler. Each feature stacked like bricks: navigation cues for the blind, itinerary tips for tourists, and a layered query system where a single image could unravel through endless questions. By the final hour, Bright Walk wasn’t just an app—it was a bridge, turning “I can’t” into “I explored.”

Tech Stack

Kotlin
Python
Java
XML

Fundraising Status

We haven't raised the funds yet , but considering the vast opportunities while catering tourist and helping visually impaired , we are in a strong position to raise money for market deployment.