Page 1 of 1

BITalino-based emotion detection platform

Posted: Wed Dec 13, 2017 4:39 pm
by pmisiak

I am a student of Computer Science at AGH University of Science and Technology in Krakow, Poland, Europe and I am preparing my MSc thesis about using biometric data to inference person's emotional state. This platform is to consist of BITalino measurement kit and an Android application with built-in emotion classifier (e.g. neural network or SVM - these are the terms from the area of machine learning).

I have found an article"Effect of Experimental Factors on the Recognition of Affective Mental States through Physiological Measures" (available at about testing different emotion classifiers using ECG, EMG and GSR signals. Such set of features is perfect for applying BITalino kit to it, however I cannot get in touch with the authors of this papers.

I would like to ask if anybody performed any experiment consisting on measuring biometric signals and combining them with emotional states. I need a classifier or a dataset which I could use to train my own classifier.

I will be grateful for any advice or help.

Best regards,
Piotr Misiak
Department of Applied Computer Science
Faculty of Electrical Engineering, Automatics, Computer Science and Biomedical Engineering
AGH University of Science and Technology
30 Mickiewicz Avenue
30-059 Krakow, Poland

Re: BITalino-based emotion detection platform

Posted: Thu Dec 14, 2017 12:57 am
by BITalino
Hi Piotr,

Firstly we'd like to wish you all the best for the work you are starting.

One resource worth looking into is the DEAP dataset:

In addition to our forum, perhaps you can also post your question on Research Gate:

Best regards,
The BITalino Team

Re: BITalino-based emotion detection platform

Posted: Mon Jan 01, 2018 7:33 pm
by pmisiak
Thank you very much for your answer and for your help.

Yes, I already know the DEAP dataset and a few more like AMIGOS, EATMINT and others. However, the data in these dataset are quite different from what my BITalino platform offers to me, that's why I try to find something more suited to my needs. What's more, I'd like to avoid reinventing the wheel hoping someone has already done such a work.

As you advised me, I asked a few questions on Research Gate and I'm waiting for the answers if I get such. Anyway, if that doesn't help, I will of course try to adapt available datasets and train my own classifier.

Best regards,
Piotr Misiak