Sentiment Analysis for Health Assessment System

This Final Year Project represents my first major full stack application combining cognitive games with facial expression recognition for cognitive health assessment. As a beginner project, it required extensive research, learning, and guidance from mentors and professors. The system features a React frontend with a cognitive game interface that records user expressions through webcam, a Django REST Framework backend that processes video frames through a deep learning model, and a five-layer CNN model trained on FER-2013 dataset using TensorFlow Keras for facial expression recognition.

Challenges

•Finding appropriate datasets for CNN neural network training - required extensive research to identify FER-2013 as a suitable dataset for facial expression recognition
•Determining the optimal number of layers for the CNN model - needed guidance and research to understand deep learning architecture best practices
•Learning deep learning concepts from scratch - this was my first experience with neural networks, requiring significant time to understand CNNs, TensorFlow, and Keras
•Integrating webcam functionality with React frontend - faced challenges with browser permissions and real-time video processing
•Connecting Django REST Framework backend with the deep learning model - required learning how to serve ML models in production environments
•Understanding facial expression recognition concepts - needed to research computer vision and emotion detection methodologies
•Managing project timeline and scope - balancing learning new technologies while delivering a complete project

Solutions

•Conducted extensive research on available datasets and selected FER-2013 dataset after reviewing multiple options and consulting with professors
•Researched CNN architectures and consulted with mentors to determine that a five-layer CNN would be appropriate for the task complexity
•Dedicated significant time to learning deep learning fundamentals through online courses, tutorials, and documentation
•Learned React webcam integration through documentation and examples, implementing proper permission handling
•Studied Django REST Framework documentation and ML model serving patterns to successfully integrate the backend
•Researched facial expression recognition papers and methodologies to understand the domain better
•Created a structured learning plan and sought guidance from professors and senior developers throughout the project

Results

•Successfully completed the Final Year Project and got approved by the evaluation panel
•Gained hands-on experience in full stack development, deep learning, and computer vision
•Learned to work with React, Django REST Framework, TensorFlow, and Keras through practical implementation
•Developed understanding of CNN architectures and neural network training processes
•Created a functional application that combines cognitive games with AI-powered facial expression analysis
•Demonstrated ability to learn complex technologies and apply them to solve real-world problems
•Received positive feedback from professors on the project concept and implementation

Sentiment Analysis for Health Assessment System

Challenges

•Finding appropriate datasets for CNN neural network training - required extensive research to identify FER-2013 as a suitable dataset for facial expression recognition

•Determining the optimal number of layers for the CNN model - needed guidance and research to understand deep learning architecture best practices

•Learning deep learning concepts from scratch - this was my first experience with neural networks, requiring significant time to understand CNNs, TensorFlow, and Keras

•Integrating webcam functionality with React frontend - faced challenges with browser permissions and real-time video processing

•Connecting Django REST Framework backend with the deep learning model - required learning how to serve ML models in production environments

•Understanding facial expression recognition concepts - needed to research computer vision and emotion detection methodologies

•Managing project timeline and scope - balancing learning new technologies while delivering a complete project

Solutions

•Conducted extensive research on available datasets and selected FER-2013 dataset after reviewing multiple options and consulting with professors

•Researched CNN architectures and consulted with mentors to determine that a five-layer CNN would be appropriate for the task complexity

•Dedicated significant time to learning deep learning fundamentals through online courses, tutorials, and documentation

•Learned React webcam integration through documentation and examples, implementing proper permission handling

•Studied Django REST Framework documentation and ML model serving patterns to successfully integrate the backend

•Researched facial expression recognition papers and methodologies to understand the domain better

•Created a structured learning plan and sought guidance from professors and senior developers throughout the project

Results

•Successfully completed the Final Year Project and got approved by the evaluation panel

•Gained hands-on experience in full stack development, deep learning, and computer vision

•Learned to work with React, Django REST Framework, TensorFlow, and Keras through practical implementation

•Developed understanding of CNN architectures and neural network training processes

•Created a functional application that combines cognitive games with AI-powered facial expression analysis

•Demonstrated ability to learn complex technologies and apply them to solve real-world problems

•Received positive feedback from professors on the project concept and implementation