Emotion-Based Face Filter

A Computer Vision tool using OpenCV for face detection and FER for emotion recognition (~70% accuracy). Applies custom filters (e.g., smiley for "happy") and displays emotion at the bottom. Inspired by Snapchat filters.

Features

• Face detection with Haar Cascade Classifier.
• Emotion recognition (happy, sad, angry) with FER.
• Emotion-specific overlays and bottom text display.

Results

• Happy face → Green circle, "Emotion: happy"
• Sad face → Blue teardrop, "Emotion: sad"

Setup

• Run in Google Colab with OpenCV and FER.
• Requires haarcascade_frontalface_default.xml.

Demo

Emotion: Happy (0.99)

Emotion: Sad (0.65)