WeaveOn: AI-Powered Child Development App
An AI-powered Android app that helps parents understand children with autism through emotion based activity analysis, awarded 1st place at the Hology Software Development competition.
Understanding the Context
Overview
The Problem
Parents of children with autism often struggle to identify their child’s interests and potential due to limited tools that provide objective and structured insights.
The Purpose
This project aims to assist parents by leveraging facial emotion recognition to analyze children’s responses during activities and provide meaningful insights into their interests and development.
My Contributions
Responsibilities
Developed the entire Android application using Jetpack Compose
Implemented authentication flow and local login persistence
Built and integrated AI-based facial emotion recognition system
Designed and developed backend API using FastAPI for AI processing
Handled end-to-end integration between mobile app, backend, and AI services
Technologies Used
Tech Stack
📱 Frontend
⚙️ Backend
🤖 AI/ML
🛠️ Tools
Core Functionality
Features
Wevy
Analyzes children’s facial expressions during activities to identify emotional responses and potential interests.
Exercise
Provides categorized activities tailored to different developmental aspects such as logic, art, motor skills, and social interaction.
Report
Displays summarized emotion analysis results in graphical form to help parents understand patterns and dominant interests.
Visual Design
UI Showcase
Interface design and user experience highlights
Onboarding screens introducing WeaveOn and guiding parents through initial setup.
Technical Journey
Challenges
Managing user authentication state to ensure users remain logged in across app sessions.
Learned to implement local storage mechanisms for session persistence in Android applications.
Building an AI system based on an existing model and adapting it for real time emotion analysis.
Gained hands-on experience modifying a facial emotion recognition model and calculating emotion results per frame.
Designing and deploying a backend API to handle AI processing.
Learned how to build and expose AI services using FastAPI and integrate them with a mobile application.
Developing the entire application independently within a limited timeframe.
Improved end-to-end development skills, from UI slicing and state management to backend and AI integration.
Results & Achievements
Impact
Measurable outcomes and project success metrics
Awarded 1st Place at Hology Software Development Competition
Successfully delivered a functional AI-powered Android application
Demonstrated end-to-end mobile, backend, and AI integration
Next Steps
Future Plans
Conduct usability testing with parents and caregivers to validate user experience
Improve emotion recognition accuracy through model tuning and dataset expansion
Add offline support for areas with limited internet connectivity
Implement user analytics to track feature usage and engagement
Expand platform support to include iOS and web dashboard for parents