Smiley Messenger: An Avatarbased Asynchronous Communication Application

1
Smiley Messenger An Avatar-based Asynchronous
Communication Application

• George Mutale
• Andrew Marriott (Supervisor)Curtin University
  of Technology

2
Smiley Messenger An Avatar-based Asynchronous
Communication Application

• Background
• Aim and Significance
• Research and Design Methodology
• Evaluation and Future work
• Conclusion
• Questions and Answers

3
Background

• Online Communication
• Emotion Affective Computing
• Embodied Conversational Agents (ECAs)
• Computer Animation
• Existing Technologies (MetaFace FaceFactory)
• Evaluation

4
Background

• Online Communication (e-mail)
• E-mail
• Asynchronous
• Miscommunication
• Smilies

5
Background

• Online chat
• Synchronous
• Avatars
• Avatars
• Online forums
• Online chat
• Massive multiplayer online roleplaying game
  (MMORPG) . E.g. Anarchy Online

6
Background

• Emotion
• No single definition
• Universal expressions
• Emotion in ECAs
• Affective Computing
• The need for computers to recognise and express
  emotion
• Different approaches
• Autonomous Agents

7
Background

• Face to Face
• An important aspect of communication
• Teleconferencing
• Online Identity
• Bandwidth requirements
• Embodied Conversation Agents (ECAs)
• Different conversation abilities
• Different uses and appearance
• ECA (face) to represent a person

8
HCA

• Make kids learn about H.C Andersons life, fairy
  tales and historical period while playing and
  having fun.

9
REA

• REA plays the role of a real estate salesperson
• She uses gaze, head movements and facial
  expressions for dialogue functions such as
  turn-taking, emphasis and greetings

10
Spiky Boy The MetaFace framework
11
Aims

• Enable a user to create engaging and affective
  content in e-mail using 2D Smilies (ECAs)
• Evaluation of the affectiveness of the received
  e-mail

12
Significance

• E-mail Creation
• A professional system that enables a user to
  script multimedia content to send as e-mail
• Evaluation
• Attempt to establish what aspects of the ECAs
  make them engaging and or affective
• Public access to the research tool and evaluation
  to provide a close to real world scenario as
  opposed to a controlled experiment

13
Hypotheses

• Smiley Messenger will be an easy messaging tool
  to use.
• Smiley Messenger will produce engaging message
  content.

14
Research Methodology

• Research Methodology
• Development and Evaluation
• Informal interviews

15
Design Methodology

• Waterfall Model (Royce 1970)

16
Architecture

• Existing Talking Head Technology used
• PHP, Java, Perl, MySQL, Web service, XML

17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
Evaluation

• Smiley Messenger
• Asynchronous Communication system
• Allowed creation and customization of messages
• Over 65 users
• Observations
• Tutorial and Information sheet

21
Smiley Messenger Compose

• A selection of
• Faces
• Voices
• Size of final animation
• Control of speech
• Sad, happy, surprised, neutral, disgusted
• Pause
• Compose page

22
Composing a Message
23
Smiley Messenger Emotion Tags
24
Smiley Messenger Viewer

• Multiple message formats
• Movie (AVI)
• GIF image with animated text
• Accessible from anywhere in the world
• Easy to compose a message

25
Engaging message content
26
Affective message content
27
Summary of other data collected
28
Summary of other data collected

• Voice needed improvement
• More face options
• Facial expressions
• GIF Text was distracting
• Thought smilies were cute and appealing
• Interesting and nice way to send a message

29
Future Work

• Smilies with voice (Smiley Central)
• Face Robot (SOFTIMAGE)
• Adding additional tags to the smiley messenger
• Add facial expressions
• Improve Audio (Text to Speech)
• Further evaluation

30
Conclusion

• Hypotheses
• Smiley Messenger will be an easy messaging tool
  to use.
• Smiley Messenger will produce engaging message
  content.
• The results of the research show that both
  hypotheses were met.
• ECA can be used as a medium for communicating
  affective, engaging content.
• Questions

31
(No Transcript)
32
Face Factory