behnaz farahi heart of the matterbehnazfarahi.com/assets/img/affective computing in... · more...

2

HEART OF THE MATTER:

1 Mesolite: An emotive display

case responding to the facial

expressions of the viewers.

Behnaz FarahiUniversity of Southern

California (USC)

Affective Computing in Fashion and Architecture

1

ABSTRACTWhat if material interfaces could adapt physically to the user’s emotional state in order to

develop a new affective interaction? By using emotional computing technologies to track

facial expressions, material interfaces can help to regulate emotions. They can serve

either as a tool for intelligence augmentation or as a means of leveraging an emphatic

relationship by developing an affective loop with users. This paper explores how color and

shape-changing can be used as an interactive design tool to convey emotional informa-

tion, and is illustrated by two projects, one at the intimate scale of fashion, and one at a

more architectural scale. By engaging with design, art, psychology, computer and material

science, this paper envisions a world where material systems can detect the emotional

e n e a e an e nfi e t em el e in e t ente int a ee a l wit

t e e a e ti e tate an in en e ial inte a ti n

TOPIC (ACADIA team will fill in) 3

‘Affective computing’ is now widely used to refer to the

‘computational modeling of emotion and implementations

of autonomous agents capable of affective process-

ing.’(Klaus, 2010) Put simply, affective computing is about

developing systems, which can recognize, interpret and

simulate human emotions through measuring physiological

responses. In fact, studies have shown that the majority

of affective communications take place non-verbally or

para-linguistically through facial expressions, gestures

an al in e ti n ( i a an e a ian )

Thanks to sensor technologies various data from the user’s

physiological or neurological responses can be captured

and processed. In a manner not dissimilar to how we

understand each other emotions through modalities of

various information, these systems can perceive cues for

any emotions. For instance, computer vision sensors can

be used to capture the bodily gestures and even facial

expressions, while biometric sensors can directly measure

physiological data such as skin temperature and galvanic

resistance and help us to better understand our emotional

state.

Although there is an ongoing debate as to whether emotions

are socially and culturally constructed or universal(2), in

t e Ame i an ant l i t a l man aime

to show that certain types of emotions are not cultur-

all e ifi an a e in a t ni e al in all wal li e

A an t it mm n em ti nal e ita e in

humanity together, then, in a way that transcends cultural

i e en e ( an ) man alle t e ll win a i

emotions’: joy, distress, anger, fear, surprise and disgust.

Through his research he attempted to argue that the facial

expressions associated with the basic emotions are innate

an ni e al ( man )

Universal emotions manifested in a physiological way

through facial expressions, can be detected and recognized

by computational systems. And if materials are enhanced

with these computational systems, how might matter

represent or simulate an emotional response accordingly?

In other words, how can we map various emotions to

a i e n e t t it an t e wa w mi t we

use techniques for detecting facial expressions to control

responsive behavior?

PROVOKING AN EMOTIONAL RESPONSE m ti n a t e w in i ate i a t m ement

about externalized behavior, about certain orchestrations

of reactions to a given cause, within a given environment’

(Damasio, 2000)

In it ei e an a ianne immel n te

INTRODUCTIONDoes matter ‘have’ emotion? Can material systems ‘recog-

nize’ emotion and ‘provoke’ certain emotional responses in

users? Can new materials be imbued with the right integra-

tion of sensing, actuation and communication so as to serve

as affective matter to detect and respond to emotions?

In the past many Western thinkers have viewed emotion

a an ta le t ati nal an intelli ent t in in ( an

) t e e a een a la e a etween ati nal an

emotional perspectives. Conventionally, computers are

considered as being rational and logical. They are also

thought to be good at accomplishing certain cognitive

tasks at which humans are not so good. Anything related to

emotion would therefore need to be dismissed or simply not

ta en e i l t e ientifi mm nit ( i a )

In the past decade this view has changed dramatically. At

last it has been widely accepted that emotional systems can

al in en e niti n i mean t at e e t in we

has both a cognitive and affective component that assigns

meanin a well a al e m ti n w et e iti e

ne ati e an i e tl in en e e a i an ni-

tions including our perception, attention, motivation and

in general our decision making capabilities. Advances in

neuroscience and psychology about understating the role

of emotion - such as the research by leading neuroscien-

tist, Antonio Damasio - have led many computer scientists

to attempt to create computers, which can understand

emotions(1).

DETECTING EMOTION: AFFECTIVE COMPUTING ‘Affective computing’ is a term coined by Rosalind Picard

in a a e at t e m te ien e n e en e in

( i a ) we e t e i in t i en i an e

t a e e n t i in tan e in an e Cl ne

invented a machine called a ‘stenograph’ for measuring

emotions. In his experiments, subjects used touch and

t ei fin e e e t e e a e en e em ti n

– anger, hate, grief, neutral, love, sexual desire, joy and

e e en e w ile e e ien in min e en e le

m i e aime t e i e ea a e i en e t at it i

possible to ‘counter a negative emotional state by inducing

a rather rapid shift into a positive one.’ In his book, Sentics:

e t e m ti n e tline i fin in a t

‘emotional perception and response at the intersection of

m i a t an mat emati ( a ) an ela ate

on the notion of ‘sentic’ forms. “The emotional char-

a te i e e e a e ifi tle m lati n t e

motor action involved which corresponds precisely to the

eman t e enti tate (Cl ne )

4

a very interesting experiment exploring how the brain

assigns various emotional characteristics and constructs

a t t a e ie e ent ( ei e an immel )

In their experiment they showed participants a short

simple animation and asked them to describe what they

saw happening. What they discovered is that many people

assigned certain characteristics, such as emotions, inten-

tional movements and goals, to simple shapes and their

associated movements even though there is no evidence

of any facial expression or even any indication of human

representation.

In alentin aiten e ma e an t e a inatin

e ati n in i e i le e iment in nt eti

l ( aiten e ) In i t t e e i e

Braitenberg envisioned simple robots that could produce

surprisingly complex - and seemingly cognitive - behaviors.

en t t e e e i le a e a le t m e a n a t n-

omously based on how their sensors and wheels are wired

together, it appears that they have a form of agency that

allows them to achieve a goal or even represent various

characteristics such as being aggressive, explorative,

passionate and so on. For instance a robot avoiding the

source of light can represent emotion of ‘fear’, not dissim-

ilar to how a bug escapes from the light in order not to be

caught.

e ne t e i e ti n in all t e e e ati n

is how to exploit the natural phenomenon of anthropo-

morphism - the process by which we attribute mental

characteristics to animated objects. It is fair to argue

that, by utilizing material movement or color change as an

interaction design tool, it is possible to use these tools for

emotional communication. Most of these material inter-

faces have been used/studied already for their visual and

haptic communication properties, but not necessarily for

em ti nal e e i n we e t i a ea e ea i

growing. As Strohmeier et al. note, ‘Recent explorations

into shape changing interfaces have begun to explore how

shapes might be used to output emotions’ (Strohmeier et al.

)

So, how could shape- and color-changing designs convey

emotions and how might this development change the

e e ien e e i n w w l t e a li ati n

e el ment in wea a le tentiall enefit t e w

suffer from an incapacity to understand emotional cues

m t ei en i nment w mi t an a ite t al

element enefit m en a in em ti nall wit t e i i-

tors? This paper uses two projects, which engage with

the notion of emotional computing at various scales, in an

attempt to answer to these questions.

FACIAL EXPRESSION TRACKINGMark Weiser’ notion of ‘ubiquitous computing’ has already

became a reality. We now live in a world where computa-

tional devices are embedded throughout the environment.

The notion of smart environments and smart gadgets are

becoming more and more part of the fabric of our lives,

from Fitbit tracking the number of calories that we burn,

to Apps tracking our sleeping patterns, to Nest thermostat,

learning from the pattern of our behavior in buildings. It

is time for computational systems to not only engage with

quantitative aspect of our life but also to create an inter-

face with our emotions. (Figure 2)

Applications of facial expression tracking systems include:

1. Studying the affective reaction of costumers for

marketing purposes to understand their satisfaction about

a certain product

2. Serving a customized media content for adver-

tising purposes

e in t e mental ealt atient lini al

psychology and healthcare

nit in a ial e n e e it e

in airports.

For example, in the commercial world, the chocolate

man a t e e e a e el e a i en e in t ei

retail outlets, which rewards you with a sample if you smile.

Heart of the Matter Behnaz Farahi

22 e i i n Ima e en in e n l a ial e e i n t a in


The intention is to enhance the in-store experience and

create a sales opportunities. Meanwhile their competitor,

Mondelez International, plays commercials based on an

a e an en e t e ete te t me ( ie )

Likewise, facial expression tracking has been implemented

in the media and entertainment industry in order to create

e e ien e a a eDan e ( ) w i all w e le t

control the movement of a virtual Michael Jackson through

their facial muscle movement.

This paper argues that facial expression tracking can be

embedded into the fabric of material systems for a number

of purposes:

1. The application of emotional computing into smart

wearables could augment emotional intelligence. It could

not only provide a better understanding of our emotional

state in an objective way but also give us clues about our

social settings.

2. The application of emotional computing into smart

environments/objects could create a more empathic and

engaging experience by establishing an affective loop with

the user.

OPALE: AN EMOTIVE SOFT ROBOTIC GARMENT w mi t l t in en e a e i n an int e en i e

mode accordingly?

i e ti n e i e t e e i n t ate e in ale

an emotive garment which can recognize and respond to

the emotional expressions of people around it. The aim

i t e el a t ti wea a le fitte wit an ele -

tro-mechanical system that controls the shape changing

behaviors mimicking the emotional expressions of

nl e ( i e )

man ai an animal a e me t e m t in i in

natural phenomena both in terms of their morphology as

well as communication purposes during social interaction.

In i e animal ale i m e a e t

fi e ti em e e in ili n w i i tle w en

the wearer is under threat.

The intention behind the material development of this

project was to control the location and orientation of hair-

like elements so that they might respond to underlying

forces not dissimilar to how hair stands up due to micro-

muscle contractions attached to hair follicles when we

experience goose bumps or piloerection. These involuntary

responses within the skin are due either to temperature

changes or the experience of emotions such as fear, sexual

a al an e itement( ) a ie e t i al a e ie

in ata le ili n et we e in ate eneat

the fur-like skin in order to generate deformations in the

te t e an a e l me e i t i ti n fi e n

the surface was based on a study of the architecture of the

human body. Data captured from an analysis of the surface

curvature of the human body and the underlying contours

of the muscles informed the location, density and height

fi e i t i ti n e intenti n wa t e a e ate t e

movement of underlying muscles by having the denser and

ale An em ti e t ti a ment m e a e n em ti n e le a n

6

la in fi e int t e lea la e t a li eet an late int t e ili ne

Data m a e at e anal i t e man in m t e l ati n en it an t e ei t fi e

C a le an e e e lat ( i t t)

a ne mati nt l i it n i tin i t len i al e wit a a le connections.

l n e fi e ll w t e nt t e at e eneat

( i e )

e in ata le e a i we e nt lle in a t m

designed electrical board attached to an Adafruit Feather

mi nt lle ( wit A A D A C te

M0 processor) capable of controlling an array of six low

we e t me i al len i ( A e ie ) i

facilitated the computational control of air pressure and

a i in ati n t t e A in ammin en i n-

ment. As a result, each of the six pneumatic soft composites

pockets were capable of providing dynamically controlled

texture patterns that could vary in speed and frequency of

an e t i a miniat e i e C a le ( am

e an ) an a e lat ( i t t e wi ) an lit i m

l me atte ( mA ) we e e ( i e )

From the perspective of interactive design, this project

looks closely at the dynamics of social interaction. We

tend to respond to people around us through our uncon-

scious facial expressions and bodily movements. When

surrounded by smiling people, we often smile back. And

when threatened, we often take on a defensive stance.

Through the logic of neuron mirroring, we mimic each

other’s emotional expressions. Likewise animals use their

skin, fur and feathers as a means of communicating. Dogs,

cats and mice bristle their fur as a mechanism of defence

a a m intimi ati n Da win wa t e fi t t e amine

the emotional signals in humans and animals in his book,

e e i n t e m ti n in an an Animal ( )

e a e t at t e wa t at ai tan n en w en

we are scared is a ‘leftover from a time when our ancestor

were completely covered in fur’ and its role was to make

t em l i e an m e intimi atin (Da win )

The challenge, then, is to develop clothing that can likewise

express emotions. For example, might it be possible for

clothing to sense aggression, and go into defensive mode

accordingly?

The challenge was to explore whether emotions expressed

in our social interactions could be represented in a

non-verbal way through the motion of a garment. Thus the

garment would become an expressive tool or apparatus

that empathizes with the onlookers. For this purpose, this

dress is equipped with a facial tracking camera that could

detect a range of facial expressions on the onlooker’s face:

a ine a ne i e an e an ne t al a

emotion detected was sent to a microcontroller (Teensy

) a a le a ti atin t e len i t ene ate a i

atte n an in ati n ee in ea ai et ( i e )

For example, if the dress were to detect an expression

of ‘surprise’ in an onlooker’s face, the wearer’s shoulder

a ea w l ta t t in ate w en an nl e e e e

‘anger’, the wearer’s shoulder and chest would start to

in ate an e ate wit a anti a e i e m ti n en

people around start to smile and demonstrate ‘happiness’

t e e w l i le tl m t t tt m ( i e

)

Although the current application of emotional computing

and soft robotics for wearables still has limitations, it opens

up exciting opportunities for shape-changing clothing in the

future for both communication and healthcare purposes.

Not only does the smart garment promise to become part of



a ial t a in ame a em e e int t e ili n e an an ete t nl e a ial e e i n ( a ine a ne i e an An e )

not match the desired expectation, leading to isolation and

rejection by others. Such a system can help them blend

more easily with others and over time learn appropriate

responses.

MESOLITE: AN EMOTIVE DISPLAYw an we eate an en a in em ti nal e e ien e

t me e i n an inte a ti e i la w an t e

life-like quality be told through materials, dynamic behavior

an l t an mati n w an e na an a a te

be used as a tool for interaction design?

Mesolite is an emotive display commissioned by adidas

designed to showcase the future concept shoe, and offer

retail and other consumer experiences. The aim is to

explore how computer vision and facial tracking technology

can be implemented in the design of a display in order to

in en e atte n ial inte a ti n an ma imi e t e

engagement with the viewers by giving animal-like qualities

to the designed display object. (Figure 11)

Inspired by the form of a soccer ball, the Mesolite sphere

n i t e a nal an enta nal m le C C

milled out of black acrylic. The sphere is equipped with

a facial tracking camera and has an irregular opening in

w i t ei late t e e i w a e ( ) In i e

by the natural formation of Mesolite crystal, the sphere is

m nte wit a li t e ( iamete ) a in

in length – lit up with various lighting effects to convey the

speed and movement of the player and soccer ball on the

10

e e i e n in t t e nl e em ti n wit a i t e dynamic behavior

t e a a at man intelli en e t it an al enefit

man e le wit a ti m w a e i fi ltie e ni in

facial expressions. People with autism might be paying

attention to what you are saying but be unable to tell if you

are happy, sad or angry. As a result their responses might

8

fiel ( i e )

The piece has a certain magic and wonder to it, similar to

w fi e ie emit li t in nat e m e e wit

in i i all a e a le D i el t e ti a li

tubes illuminate creating a mesmerizing dance of light. For

this purpose, from inside, every hexagon and pentagon has

a e i ate anel w i an e atta e wit et

mall ma net an w i e t e D i el a

panel connects with a female-male latching connector to

t e nei in m le All t e wi e t at nne t t e D

meet each at the bottom of the sphere where they connect

to the dedicated microcontrollers. Inside the sphere there

i lat m wit an ani li e lan a e fi e ti

( imila t ale) e i e wit D i el an an

aluminum shaft connected to a stepper motor in the middle.

The product – the adidas concept shoe - is mounted on the

a t a e a lan a e fi e ti eatin an ani

inte a ti n etween t e t e e an t e fi e ti

landscape. ( i e )

The Mesolite display has an ‘eye’, a facial tracking camera

w i an ete t t e a ial e e i n t i it

and the locations of their heads relative to the opening of

the sphere. When a face is detected, Mesolite comes out

of ‘dream mode’ and acknowledges the viewer's presence.

‘Dream mode’ refers to when the shoe is not moving, and

w en t e D atte n li t a e a tle m ement wit

a w ite limme in li t e e t n e t e a e t e i it

is detected the Mesolite comes alive, acknowledging the

presence of the viewer by generating a red ripple of light

which goes across the surface from the opening to the back

of sphere as though it is welcoming the visitor’s presence.

The shoe inside also comes alive by tracking the head of

the visitor thereby creating an intimate engagement with

the viewer. For this purpose, the <x,y> value related to the

head location of the viewers captured from the camera is

mapped to the rotational position of the shoe, giving the illu-

sion of the shoe facing the viewers. In this way Mesolite is

given a form of attention directionality. If there are multiple

people in front of Mesolite, the camera computes the values

of the detected faces all together, and the shoes start

looking at multiple faces as though it is trying to capture the

attenti n m lti le e le n e a e n a t e atten-

tion of Mesolite, she has to try to keep it by getting closer. If

not, the shoe starts switching its attention to as many faces

a ete te in nt t e enin ( i e )

The more the viewer engages, the more Mesolite comes to

life. When the viewers express surprise, the red lighting

starts to ripple in and out with deep breath-like rhythms.

When the views smile and express happiness, Mesolite will

share the happiness by having the shoe spin around and

t e e li t ta t t a a i l a ew time a t it i

also excited and happy!

The computing system for the piece includes a central brain

(Raspberry pi) and four micro-controllers which receive

information from it. The task of the Pi is to process data

from and send data to microcontrollers that interact with

t e i al w l ( ) A i ntain t e main a li-

cation loop, which is orchestrated to run approximately

11 Mesolite, an interactive display

showcasing the latest concept

shoe, responds to the excitement of

the visitors.

12 Diagram of placement of each

element in t e e e n t e i t hand side from top to bottom:

Camera, shoe, platform with

em e e D an fi e ti computational brain

C m tati nal ain in l e ( ) w we lie ( A ) ( )

Three microcontrollers (Teensy

) ( ) t ( e a in C nt lle ) ( ) a e i

i t i a a in t e i la

e a e l ati n an em ti nal facial expressions of the viewer

including surprise and happiness

can be detected. Mosilte comes

alive when a face is detected.

11



12

time a e n mm ni atin wit t e nne te

microcontrollers to obtain camera data and drive the motor

an D ne t e allen e in t e ammin

the lighting system for this piece was to map the irregular

t t e D i el t ete t t e e a t l ati n ea

i el wit t ei e i ate ID in t e D e t a e t e

sphere. To achieve this, we exported the coordinates of

ea int m t e D file an t e t em a a te t file n

t e i i t ta tin t e l i a e t e file elate

t t e inate tem t e D t tain t e

inate ea in i i al D nit ee een

mi nt lle we e e tw t em atta e t

D i el an ne atta e t t e ame a A e in

motor with its dedicated microcontroller was used to

control the shoe movements, which required the devel-

opment of our own customized software to calibrate the

ID t e m t ( ) i wa an e ential te in n e -

standing the motor torque, position and acceleration, which

was the key factor in designing the behavior.

The scope and application of this project suggests

numerous design opportunities for the world of archi-

tecture and product design. What is clear is that in

designing smart/robotic objects with life-like behaviors,

these artifacts may deliberately exploit the divergence

between the object’s characteristics and preference and

the human frame of reference. Anthropomorphism in this

context refers to the emergence of interaction between a

e an t e ti en i nment A in t le et

al., this includes emotional states, motivations, and inten-

ti n a i e t e e t t e t ( le et al )

we e t e ta e e in a i em ti n ia

shape changing interfaces, as also noted by Strohmeier et

al i a i nifi ant e i n allen e w i e i e inte -

disciplinary approach uniting design, material science

en inee in m te ien e an CI net ele

it opens up radically new opportunities for addressing

psycho-social issues.

CONCLUSIONThis paper has sought to illustrate how materials

augmented with computational tools can serve as an

emotional interface. In this process two challenges have

een a e e ne i w t ete t em ti n m t e

user and the other is how to provoke a certain emotional

response in the user through the implementation of

dynamic behaviors such as color and shape change. So,

in te m t e fi t allen e t e al i t ete t t e

viewer’s emotion using a computer vision system and to

store the information in the material system. In terms of the

second challenge, the material then responds physically

to the detected emotion in order to establish an affective

loop with users. The intention is to explore how shape- and

color-changing interfaces might be used in the future to

express and simulate various emotions through non-human

representation. These material interfaces could be a very

effective tool for the communication of emotions.

This paper has presented the design process behind two

e t ale an em ti e t ti e an e lite

an emotive display. The work described here serves as a

proof of concept for the application of affective computing

10

A i C C millin i e le m le inte nne ti n

A em l m le in tw fi in n e e e e

A em l a e i ate D anel nne te wit em e e ma net

e tin t e D li tin

in design which can be used as a tool for emotional regu-

lation either to augment emotional intelligence (e.g. with a

smart garment) or to develop an emotional bond by devel-

oping an affective loop with the users (e.g. with a smart

object display).

The intention has been to demonstrate that despite our

anxieties about smart environments and technologies in

general, there is potential for empathy where these objects

can become companions and co-exist with us and not

against us. Moreover, while the capacity of computer vision

to recognize different facial expressions has been exploited

by many commercial and advertising purposes, the inte-

gration of such a system into clothing or architectural

installation is a new venture which could open up novel

t nitie t e w l e i n CI an a ite t e

ACKNOWLEDGEMENTSThis research is part of a broader ongoing collaboration with Paolo

Salvagione and Julian Ceipek. I would like to thank them for their

advice and helpful contributions to the production of these works.

Special thank to Leon Imas, adidas for generous support during the

'Mesolite' project. Also I would like to thank to the USC Bridge Art +

ien e Allian e e ea ant am t at n e ale

NOTES1. Note that there are many different theories for emotions

borrowing from various disciplines including psychology,

neuroscience, physiology and cognitive science.

2. in tan e in e late t w m ti n A e a e i a

Feldman Barrett argues that emotions are socially and cultur-

ally constructed.

De i ne filmma e A iel lman an en t an

creative coders Aaron Meyers, Lauren McCarthy, and James

George

The fabrication process for this project consisted of manu-

all in e tin fi e ti int t e la e t m ntin

a e ( lea a li eet) A te la in all t e fi e

int t e a e t e fi e we e a e ll m e t a at

ili ne ( e ) A te n e t e ili ne wa ll

e t e m ntin a e wa em e entl m t e fi e

landscape.

a e a n an enta n wa mille ( in an A i

milling tool) using 1” thick black acrylic sheet with a designed

indent allowing each module to be connected to another module

with two sets of screws on every edge. The attempt was to have



a modular system both for ease of assembly as well as fabrica-

ti n t a m le a a t ni e an le le int

which acrylic tubes are mounted.

e ain t e i i w itten in t n wit me i e -

man e C t n e e e e t e D animati n w ile

the microcontroller code is written in a subset of C++, using

standard Ardunio and Teensyduino libraries. The code for each

an e m ile an e l e in t e A in ID ( )

A proportional–integral–derivative controller (PID controller) is

a control feedback loop mechanism mostly used for industrial

control systems. We have developed a user-friendly software

visualizer for PID controller.

REFERENCESaiten e alentin e i le e iment in nt eti

Psychology. Reprint edition edition. Cambridge, Mass.: A Bradford

Book.

Cl ne an e enti e t e m ti n e i e

edition. Bridport, Dorset: Prism Pr Ltd.

Dama i Ant ni e eelin at a en an

m ti n in t e a in C n i ne e iti n an Die CA

Mariner Books.

Da win C a le e e i n m ti n in an an

Animal n n D e inte t

man a l an i a a a e a e an

e i n ni e al m ti n in a ew inea illa e ew

York: Garland STPM Press.

an D lan m ti n A e t Int ti n I e iti n

ew ni e it e

ei e immel An e imental t A a ent

e a i e Ame i an nal l ( ) D I

la e e m ti n an m ti nal C m eten e n e -

tual and theoretical issues for modeling agents in A Blueprint for

Affective Computing: A Sourcebook and Manual. Scherer, Klaus R.,

an a an i e an tienne e e iti n ew

ni e it e

Mehrabian, A., “Silent Messages”, Wadsworth, Belmont, California,

le A a t A ali an Ca i en we

need a human: Motivational determinants of anthropomorphism,”

ial C niti n l n A

ie li e mile an e e e e in t e

excitement with facial recognition sampler. Access from

tt www n e ti ne new m A ti le

e e mile am le a ial e niti n t i en e

late tm e i t tm me i m n ite tm

campaign=copyright

Paul Strohmeier, Juan Pablo Carrascal, Bernard Cheng, Margaret

e an an el e te aal An al ati n a e C an e

C n e in m ti n In C I ee in t e C I

C n e en e n man a t in C m tin tem

i a wa A ent t at e ni e m ti n I A I A

A te ee in na

i a alin A e ti e C m tin I e ia

a at e e t al C m tin e ti n e ni al e t

i a alin e mi e A e ti e C m tin in

e an A e ti e C m tin ite a ael

Cal i ne D ell nat an at an A i a a

ni e it e

a a ia enti m ti nal ealin

Music and Touch Access from https://www.brainpickings.

man e l ne enti al Cl ne le t e n

enti C le tt www t e m wat I l

IMAGE CREDITSi e lena li

i e e na a a i

i e i la Cam ie le mit e

i e a ine i t ie

Behnaz Farahi is a designer exploring the potential of interactive

environments and their relationship to the human body working

in the intersection of architecture, fashion, and interaction

design. She also is an Annenberg Fellow and PhD candidate

in Interdisciplinary Media Arts and Practice at USC School of

Cinematic Arts. She has an Undergraduate and two Masters

degrees in Architecture.

behnaz farahi heart of the matterbehnazfarahi.com/assets/img/affective computing in... · more...

Documents