Lecture 1 - Introduction Flashcards
Artificial Intelligence
Computer Vision + Planning and search + Knowledge Representation + Machine Learning + NLP + Robotics
Social AI
Computer vision + planning & search + Knowledge representation + machine learning + NLP + Robotics + Human + Memory + Attention
Social Robot Characteristics (8)
- Express and recognize emotions
- Communicate verbally and non-verbally
- Have a personality – *Should a robot have a personality?*
- Learn social skills – *No way near here and make this list maybe for future*
- Get-to-know and recognize someone
- Maintain social relationship – *How to do this with a machine?*
- Be transparent
- Be useful
Social Interaction is?
Engangement:
- Attention
- Interest
- Affect
Co-regulation: Doing something together.
- Coordination
- Agency
- Competence
Social Awareness
Social robots must demonstrate it. Perceive, interpret and respond appropriately to verbal and non verbal cues.
Uncanny valley
Closer the robot is to human appearance the higher the rejection
Robot Interaction Design
Type of interaction Design where the type of interaction is a Social Robot
PACT
- People: stakeholders, personas
- Activities: Scenario, Specification
- Context
- Technologies
Persona
Realistic description of a typical user
Persona Components
- Persona name - Set of unique goals related to application
- User behavior, attitudes, activities, environment
- Credible details that help designers see personas as real potential user for which they can design.
Why is a persona useful?
- Help designers make design decisions
- To remind that real people will be interacting with robot
Why should we accept social robots as interfaces
- Motivation
- Ease of use
- Comfort
Scenario
- Describes human task and activities in a story
- Uses the vocabulary and phrasing of users that can be understood by stakeholders.
Problem Scenario
Story about current, existing activities
Design Scenario
story about how a robot can be used
Scenario is useful for?
help designer explore context, user needs, goals, requirements and usability issues.
Robot Dialogue Behavior
- Fully voice-controlled interaction
- Repeated statements. e.g. “let’s proceed to the next question”, were randomized from a predefined set.
- Dialogue of the robot was supported by neutral gestures performed by the robot.
- While listening the robot did no make gestures but was in breathing mode.
- The robot at all time tracked the face of the patient and tried to keep eye contact.
Reliability of Data Collection
Robot data collection overall is a reliable method, however sometimes follow up studies are needed