Affective Computing

Affective Computing PhD Course 2008/2009 Ana Paiva IST/INESC-ID

Structure of the course

• Motivation.

– What is Affective Computing? – Applications and Problems

• Perspectives on Emotions – – – –

History of Affective Sciences Communication of Emotions Emotions and Neuroscience Appraisal Theories

• The Affective Computer

– Architecture and components

• Recognising Emotions and other Affective states – Facial and Gesture Recognition – Physiological signals – Speech recognition

• Emotion Synthesis • The expression of emotions in computers and

Bibliography • Affective Computing, Rosalind Picard, MIT Press, 1997 • Understanding Emotions, K. Oatley, D. Keltner & J. Jenkins, Blackwell Publishing, 2006

Evaluation • Preparation and presentation of one seminar on a specific topic (20%) • Readings, summaries and discussion done during the classes (20%) • Project on Affective Computing (60%)

Affective Computing Computing (rational) Affect (non rational)

Affective Computing: a contradiction?

Motivation • In Artificial Intelligence: In 1967 Herb Simon emphasised that a general theory of thinking and problem solving needs to consider and incorporate the influences of emotion. • In HCI: The communication with the machines needs to take into account the affective state of the user (captured by his/her voice, facial expressions, and other signals) • In HCI and AI: the expression of computers when interacting with users must be natural and inspired in the way humans communicate, thus emotions should be present in such communication.

What is Affective Computing? “Computing that relates to, arises from, or deliberately influences emotions” Rosalind Picard, 1997

What is and What is Not Affective Computing?

Emotion and Affect • What are emotions? • What causes them? • Why do we have them?

The Limbic System The Limbic System is the seat of emotion, memory and attention. It helps determine “valence”(if you feel positive or negative) and salience (what gets the

The Limbic System • LeDoux findings show that the audio cortex is not always needed for auditory conditioning • The amygdala is where the learning for fear conditioning occurs. • There are substantially more connections from the limbic system to the cortex than vice versa. • “Not only can the limbic system hijack the cortex, but the limbic system influence

Decision Making and Emotions • “Rational thinking” requires participation from emotionmediating parts of the brain. • The importance of Emotional Inteligence as “the ability to monitor one’s own and others feeling and emotions, to descriminate among them and to use this information to guide one’s thinking and action” (Gardner).

Affective Communication • “Basic affect recognition and expression are expected by humans in communication” • So far, computer-based communication is “affect-blind”, “affect-deaf”, and generaly speaking affect-impaird.

A quantum leap in communication will occur when computers become able to at least recognise and express affect.

• • • • • •

Comunication of Emotions Emotions and Neuroscience (Sergio) Emotion Recognitionn (Pedro) Emotion Agent Architectures (Samuel) Expression of Emotions (Iolanda) Applications of Affective Computing (André) • Affective Wearables (Rui)