Audience response analysis system based on emotion recognition algorithms

⌨ team project for [2019-2] KNU capstone project2
🖱 team mentor justinjoy 🎥 Demo Video: https://www.youtube.com/watch?v=6N2loif-hoY&feature=youtu.be

Introduction

A system that detects and analyzes people's faces in videos using Amazon Rekognition API Statistical analysis results to quantify audience responses

System Architecture

Results

facial expression-based appraisal system

1. Get snapshot while taking a video(GStreamer/cv2)
2. save snapshot in local to web storage
3. emotion analysis of snpashot
4. process the anlaysis results into "emotion formula"
5. Visualize Results

"emotion formula"

we formulate a formula with real video and people's reaction