NgeMood

Description

NgeMood is a web-based application designed to help users understand their emotional conditions through facial analysis and text-based Artificial Intelligence.

This platform integrates computer vision, natural language processing, and generative AI into a unified system to present a personal, interactive, and data-driven self-reflection experience.

Background

Mental health issues are becoming increasingly relevant, especially among students and young generations. However, many individuals still struggle to recognize, understand, and reflect on their emotional conditions consistently.

Most emotions are only felt momentarily without any recording or further analysis, making it difficult to understand long-term emotional patterns.

NgeMood was developed as a technology-based solution to help users build better self-awareness by leveraging AI in analyzing facial expressions and journal texts, providing contextual and easy-to-understand insights.

Key Features

• Facial emotion detection using deep learning models (facial emotion recognition)

• Sentiment analysis from journal texts using AI

• Interactive journaling with contextual and relatable AI responses

• Mood history and emotional trend visualization

• Activity recommendations based on emotional conditions

• User authentication and management system

• Interactive dashboard for monitoring emotional conditions

Roles & Contributions

• Developing the system end-to-end (Machine Learning, Backend, and Frontend)

• Training and integrating AI models for facial emotion detection using TensorFlow

• Implementing computer vision pipelines (face detection, preprocessing, inference)

• Developing backend API using FastAPI for handling authentication, data, and AI inference

• Integrating generative AI (LLM) for journal text analysis and contextual response generation

• Building an interactive frontend using Next.js with a focus on user experience

• Designing an integrated system architecture between AI, backend, and frontend

• Managing data flow from user input (camera & text) to analysis results in real-time

• Optimizing system performance to ensure low latency during inference

Impact & Results

• Presenting a digital solution to enhance emotional awareness based on AI

• Helping users understand emotional patterns through historical data and automatic analysis

• Combining AI technology and human-centered approaches in one platform

• Implementing a real-world end-to-end AI system in the mental health domain