Blog 01: Affective Computing
The first week of this course was very intriguing as I was introduced to the composition of concepts that lie under the branch of affective computing. Previously, I’ve had a vague idea of affective computing but first week’s readings offered me insights regarding the array of subjects aligned in this area of study. Topics like emotion recognition, developing and learning from emotion, cognitive aspects of emotion have made me curious to dive deeper into this field. Affective computing is a field of studying the number of ways and aspects which can be used by a computer to help understand humans in order to implement interaction. Emotion is the prime concept around which the subject affective computing is based. A number of things come to mind when we read the word emotion. These things may differ from person to person and also may be subject to the current mood of the person. So as it stands, there could arguably be a number of definitions for the word emotion but also no definition being right or wrong.
The field of affective computing is inclusive of different techniques to identify human emotions like looking at your face, listening to your voice, noticing your gestures, and appraising the situation you are in. An introduction to these topics in the readings hooked my interest further in this field. The readings also introduced me to the term sentic modulation, which is known to be a natural and usually subconscious way of expressing emotions. As said by Beethoven in one of the readings "The face is where our eyes linger during a conversation.” - this statement generated a sense of realization in me regarding sentic modulation techniques which if combined with facial expressions can result in obtaining high accuracy for affective computation.
Moving on, another aspect of human cognition which caught my interest was the evaluation of usage of social display rules to understand emotions. This was something I never thought about before. Further views after learning about this can be quoted as “An aspect which is inherited by all humans and cannot be faked.” I further understood that collection and analysis of this data can be effectively used to understand user context and enable computers to respond appropriately creating more value for the system and resulting systems to be more intelligent as well as interactive.
Along with the introduction to the field and different theories, the readings also put light on different aspects of emotion. Emotions are categorized as physical and cognitive. The knowledge of how these aspects combined can contribute in making computers able to recognize, express and have emotions was gained through the readings. Moreover, the fact of how mapping plays a crucial role for implementers to effectively develop emotions in computers is also one of my major learnings this week.
I was honestly surprised to know the time it has taken for the research in this area to actually come into existence in the form of personal assistants. Today, we have agents that can sense emotions and are impacting the fields of automation, healthcare, education, media and communications, entertainment and many more. The readings also give insights to the fact that how difficult it was in the past for computers to collect data about human cognition considering the long screen time required to observe humans. But as it stands, we humans now on average spend around 8hours in front of a screen daily resulting in a huge scope for future research and development.
As I come to the end of summarizing my learnings from week 1, the acute introduction about this field has raised a lot of questions in my head. These questions make me curious about topics like privacy concerns, consent regulations as well as data practices around this field. But as this is just the start and there’s still a lot to learn, I look forward to finding answers to my questions in the coming weeks. That’s it from my side for week1, stay tuned to this space for my journey with the course Affective Computing!
References:
[1] Picard – Affective Computing Ch.1: Emotions are physical and cognitive Picard – Affective Computing Ch.1: Emotions are physical and cognitive - Alternative Formats
[2] Brave & Nass – Emotion in HCI Brave & Nass – Emotion in HCI - Alternative Formats
[3] McDuff & Czerwinski – Designing emotionally sentient agents McDuff & Czerwinski – Designing emotionally sentient agents - Alternative Formats