Personal tools
You are here: Home → Toolbox → SAL Database (Sensitive Artificial Listener)
Log in


Forgot your password?
 
Views
Document Actions

SAL Database (Sensitive Artificial Listener)

Overview
Description:

A database of induced emotion using an interactive artificial personality. The data is labelled for emotion using Feeltrace. (See Toolbox Labelling Tool)

Contact:

c.cox@qub.ac.uk
Language:

English
Website:

http://www.image.ntua.gr/ermis

Papers:

IST-2000-29319, D09

Access:

Data available on CD. Contact Cate Cox: c.cox@qub.ac.uk

Fact Sheet
Categorized by modality:
AV
Ethical Clearance Issues:

Release permission sought from all participants
Design/Structure:

Dataset with data analysis
Purpose of collection:

To induce and collect samples of emotionally coloured discourse.
Format:

.mpeg files, .wav files
Elicitation method:

Induced: subjects talk to artificial listener & emotional states are changed by interaction with different personalities of the listener
Nature of material:

Interactive discourse, Subjects unscripted, Machine scripted,Pilot Study 20 subjects; 4 subjects x2sessions
Size:

Approx 10 hours
Type of emotional description:

Dimensional labelling, Categorical labelling
How was emotion determined?:

FEELTRACE (See Toolbox-labelling tool for details)
Emotional content:

Wide range of emotions/emotion related states but not very intense.
 
Powered by Plone

Portal usage statistics