Paper Link: Multi-class Emotion Classification Using EEG Signals
YouTube link: YouTube
Emotion recognition can be achieved by obtaining signals from the brain by EEG. This test records the activity of the brain in form of waves. We have used DEAP dataset on which we are classifying the emotion as valance, like, arousal, dominance. We have used LSTM and CNN classifier which gives 88.60% and 87.72% mean accuracies respectively (Better/Comparable to existing models) to predict the model successfully.
DEAP dataset can be found on the official website.
Mentor: Divya Acharya
Organisation: LeadingIndiaAI
Duration: Jul 2020 - Aug 2020 (6 weeks)
Team Members:
- Riddhi Jain
- Siddhi Jain
- Rahul Sahni
- Sanika Deshmukh
- Siba Smarak Panigrahi, third-year undergraduate student at the Department of Computer Science and Engineering, IIT Kharagpur