Facial Expression Analysis Using Haar-Cascade Face Detection and Convolutional Neural Networks

Facial Expression Analysis Using Haar-Cascade Face Detection and Convolutional Neural Networks

Sindhushree P V,Sudha kumari,Yashaswini G S,Mrs. Sushma B A,Dr Krishna Kumar P R,Mrs Jayashri M

Dept of CSE
SEA College of Engineering & Technology

Abstract

Facial expression analysis is a crucial aspect of human-computer interaction, enabling machines to recognize and respond to human emotions effectively. This project presents a system that combines Haar-Cascade face detection with Convolutional Neural Networks (CNNs) to identify and classify facial expressions in real-time. The Haar-Cascade classifier is utilized for efficient and accurate face localization in static images or video frames. Once the face is detected, it is processed and fed into a trained CNN model that classifies the expression into predefined categories such as happy, sad, angry, surprised, neutral, etc. The CNN architecture is designed to extract hierarchical features from facial images, capturing subtle variations in muscle movement and facial landmarks. The system is trained and evaluated on benchmark facial expression datasets, demonstrating high accuracy and robustness under various lighting and background conditions. This integrated approach offers a reliable and real-time solution for applications in emotion-aware systems, surveillance, healthcare, and human behavior analysis.