ClearSight AI: Assistive App for Blind and Visually Impaired People
ClearSight AI: Assistive App for Blind and Visually Impaired People
ClearSight AI is an assistive app that helps blind and visually impaired users recognize objects through real-time audio feedback.
ClearSight AI is an assistive app that helps blind and visually impaired users recognize objects through real-time audio feedback.
Problem
The problem I’m trying to solve is that blind and visually impaired people often have trouble identifying objects around them in their everyday lives. My idea is to create an app that uses a phone’s camera to recognize objects and tell the user what they are through voice feedback. This will help users be more independent in their daily tasks, like navigating a room, finding items, or identifying people nearby. The app will also support regional languages to make it accessible to a wide range of users.
The problem I’m trying to solve is that blind and visually impaired people often have trouble identifying objects around them in their everyday lives. My idea is to create an app that uses a phone’s camera to recognize objects and tell the user what they are through voice feedback. This will help users be more independent in their daily tasks, like navigating a room, finding items, or identifying people nearby. The app will also support regional languages to make it accessible to a wide range of users.
Target Audience
Primary Audience: Blind or visually impaired people of all ages.
Secondary Audience: Caregivers, friends, and family members who want to help visually impaired individuals become more independent.
Primary Audience: Blind or visually impaired people of all ages.
Secondary Audience: Caregivers, friends, and family members who want to help visually impaired individuals become more independent.
Key Goals
➡️ Make it easier for blind people to identify objects in their environment through real-time voice feedback.
➡️ Provide a simple interface that does not require the user to see or touch the screen.
➡️ Support multiple languages, including regional languages, so anyone can use the app in their native language.
➡️ Make it easier for blind people to identify objects in their environment through real-time voice feedback.
➡️ Provide a simple interface that does not require the user to see or touch the screen.
➡️ Support multiple languages, including regional languages, so anyone can use the app in their native language.
Possible Solution
To help blind and visually impaired individuals identify objects around them, this app will use a phone’s camera to recognise objects and provide real-time voice feedback. This app empowers users to navigate their surroundings, find items, or recognise people independently.
To help blind and visually impaired individuals identify objects around them, this app will use a phone’s camera to recognise objects and provide real-time voice feedback. This app empowers users to navigate their surroundings, find items, or recognise people independently.
Features of the App
➡️ Volume Button Control
➡️ Object Recognition via Camera
➡️ Voice Feedback in Multiple Languages
➡️ Offline Mode
➡️ Volume Button Control
➡️ Object Recognition via Camera
➡️ Voice Feedback in Multiple Languages
➡️ Offline Mode
Design Thinking Process
Empathize
The first step was to deeply understand the challenges visually impaired people face daily.
User Interviews: I spoke with visually impaired individuals to understand their struggles with identifying objects, navigating spaces, and using current accessibility tools.
Observations: I spent time observing how they interact with objects and their environment to identify pain points.
Insights:
They often find it difficult to identify objects like keys or phones.
Reliance on others for everyday tasks such as finding things or moving through unfamiliar spaces is common.
Complex apps with small buttons are frustrating to use.
Empathize
Based on the insights from the empathy stage, I clearly defined the problem.
Problem Statement:
Visually impaired individuals have trouble identifying objects around them and navigating their environment, which impacts their independence and requires reliance on others for assistance.Key Insights:
There’s a need for simple, voice-controlled object recognition.
Hands-free, quick interaction is crucial for ease of use.
Supporting regional languages is essential for accessibility.
Ideate
Next, I brainstormed possible solutions to address these challenges.
Idea 1: Use the phone’s camera to recognize objects and give audio descriptions.
Idea 2: Make the app accessible with volume button controls, so users don’t need to search for on-screen buttons.
Idea 3: Offer regional language support for users in different parts of the country.
Idea 4: Use voice commands to let users easily switch between scanning and settings.
Prototype
I then created a basic version of the app to test its core features.
App Interface: It has a simple layout, but everything is voice-driven. Users can start object scanning with a press of the volume button.
Core Features:
Object recognition through the camera with voice feedback.
Easy access via volume button control and voice commands.
Offline mode for when internet connectivity is not available.
Support for regional languages like Marathi or Hindi for wider accessibility.
Test
As this is still an idea in the concept stage, I haven’t yet developed the product, so user testing with visually impaired individuals is pending.
Iteration
Since the product hasn’t been fully developed, the iteration process based on user feedback will be completed once testing is done.
Empathize
The first step was to deeply understand the challenges visually impaired people face daily.
User Interviews: I spoke with visually impaired individuals to understand their struggles with identifying objects, navigating spaces, and using current accessibility tools.
Observations: I spent time observing how they interact with objects and their environment to identify pain points.
Insights:
They often find it difficult to identify objects like keys or phones.
Reliance on others for everyday tasks such as finding things or moving through unfamiliar spaces is common.
Complex apps with small buttons are frustrating to use.
Empathize
Based on the insights from the empathy stage, I clearly defined the problem.
Problem Statement:
Visually impaired individuals have trouble identifying objects around them and navigating their environment, which impacts their independence and requires reliance on others for assistance.Key Insights:
There’s a need for simple, voice-controlled object recognition.
Hands-free, quick interaction is crucial for ease of use.
Supporting regional languages is essential for accessibility.
Ideate
Next, I brainstormed possible solutions to address these challenges.
Idea 1: Use the phone’s camera to recognize objects and give audio descriptions.
Idea 2: Make the app accessible with volume button controls, so users don’t need to search for on-screen buttons.
Idea 3: Offer regional language support for users in different parts of the country.
Idea 4: Use voice commands to let users easily switch between scanning and settings.
Prototype
I then created a basic version of the app to test its core features.
App Interface: It has a simple layout, but everything is voice-driven. Users can start object scanning with a press of the volume button.
Core Features:
Object recognition through the camera with voice feedback.
Easy access via volume button control and voice commands.
Offline mode for when internet connectivity is not available.
Support for regional languages like Marathi or Hindi for wider accessibility.
Test
As this is still an idea in the concept stage, I haven’t yet developed the product, so user testing with visually impaired individuals is pending.
Iteration
Since the product hasn’t been fully developed, the iteration process based on user feedback will be completed once testing is done.
User Persona
Information Architecture
1. Home Screen (Main Functionality)
├── Welcome Audio (Intro message)
├── Start Scanning (via Volume Button)
├── Motion Feedback (Vibration once Image Scan Done)
| ├── Audio Feedback (Object descriptions)
| └── Voice Command: "Start Scan"
└── Stop Scanning (via Volume Button)
├── Audio Feedback (Stop confirmation)
└── Voice Command: "Stop Scan"
2. Settings
├── Language Settings
| ├── Select Language (e.g., Marathi, Hindi)
| └── Voice Command: "Change Language"
├── Voice Settings
| ├── Adjust Speed, Pitch, Gender
| └── Voice Command: "Adjust Voice Settings"
├── Offline Mode
| ├── Toggle On/Off
| └── Voice Command: "Enable/Disable Offline Mode"
├── Accessibility Features
| ├── Vibration Cues (On/Off)
| └── Voice Command: "Turn Vibration On/Off"
└── Tutorial
├── Replay Onboarding Instructions
└── Voice Command: "Open Tutorial"
3. Help Section
└── Help and Support (Audio guides)
└── Voice Command: "Help"
4. Exit/Close App
└── Confirmation Audio Prompt
└── Voice Command: "Close App"
On-Boarding Flow - Final Visual Design
Main Functionality - Final Visual Design
Feasibility and Conclusion
This App is technically feasible to develop using existing technologies and APIs. The key components—object recognition, voice feedback, multi-language support, offline mode, and volume button control—are all achievable using widely available APIs like Google Cloud Vision, Google Text-to-Speech, and TensorFlow Lite.