Commit 6cc97869 authored by Lukas Hindemith's avatar Lukas Hindemith
Browse files

Add media files and add presentation slides.

parent 8c36c036
......@@ -2,6 +2,11 @@ span.highlight{
color:orangered;
}
div.citations{
line-height: .4em;
margin-top: 1em;
}
span.citation{
font-size: 0.3em;
font-style: italic;
......@@ -10,4 +15,8 @@ span.citation{
table th.no-border,
table td.no-border{
border: none;
}
.side_by_side{
display: flex;
}
\ No newline at end of file
......@@ -37,6 +37,7 @@
<li>Improving Mental Models I</li>
<li>Robot Control Architecture</li>
<li>Improving Mental Models II</li>
<li>Perspectives</li>
</ul>
</section>
<section>
......@@ -73,23 +74,309 @@
<span class="citation">Honig, S., & Oron-Gilad, T. (2018). Understanding and resolving failures in human-robot interaction: Literature review and model development. Frontiers in psychology, 9, 861.</span>
</section>
<section>
<h2>Overall Approach</h2>
<ul>
<li>Improve Mental Model</li>
<li>Communicate technical abilities and limitations of robot</li>
</ul>
</section>
<section>
<h2>Research Questions</h2>
<ol>
<li class="fragment">Mental Model Improvement<br>
<span style="font-size: 0.5em; font-style: italic;">How can naive users' mental models of robots be improved?</span>
</li>
<li class="fragment">Concept Realization<br>
<span style="font-size: 0.5em; font-style: italic;">How should technical concepts of robots' architecture be realized to be understandable
for naive users?</span>
</li>
<li class="fragment">Interaction Improvement<br>
<span style="font-size: 0.5em; font-style: italic;">Does users' knowledge about the robot's concepts improve Human-Robot Interaction?</span>
</li>
</ol>
</section>
</section>
<section>
<section>Scientific Background</section>
<section>
<h2 class="r-fit-text">Scientific Background</h2>
</section>
<section>
<h2>Mental Models</h2>
<div class="r-stack">
<span class="fragment fade-in-then-out citation">James K Doyle and David N Ford. “Mental models concepts for system dynamics research”. In: System dynamics review: the journal of the System Dynamics Society 14.1 (1998), pp. 3-29.</span>
<img class="fragment fade-in-then-out r-stretch" src="media/mm/mental_model_1.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/mm/mental_model_2.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/mm/mental_model_3.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/mm/mental_model_4.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/mm/mental_model_5.svg">
</div>
</section>
<section>
<h2>Implications for HRI</h2>
<h4>Expectations Towards Robots</h4>
<ul>
<li>Influencing initial mental model</li>
<li>Previous experiences</li>
<li>General expectations</li>
</ul>
</section>
<section>
<h2>Implications for HRI</h2>
<h4>Media Equation & Anthropomorphization</h4>
<ul>
<li>Media Equation:
<ul>
<li>eliciting emotions through media</li>
<li>Applying rules of social conduct towards computers</li>
</ul>
</li>
<li>Anthropomorphization
<ul>
<li>Attribution of humanlike properties, characteristics or mental states to nonhuman agents and objects</li>
</ul>
</li>
<li>misguided perceived structure of the system -> wrong mental model</li>
</ul>
</section>
<section>
<h2>Implications for HRI</h2>
<h4>New Directions of Robot Design</h4>
<ul>
<li>Otherware</li>
<li>Techno mimesis</li>
</ul>
</section>
<section>
<h2>Robot Transparency</h2>
<h4>Social Feedback</h4>
<ul>
<li></li>
</ul>
</section>
<section>
<h2>Robot Transparency</h2>
<h4>Technical Feedback</h4>
<ul>
<li></li>
</ul>
</section>
</section>
<section>
<section>Ways of improving Mental Models</section>
<section>Motivation and introduction of "introduction" and "visualization</section>
<section>Study Setup</section>
<section>Results</section>
<section>Limitations</section>
<section>
<h2 class="r-fit-text">RQ1: How can the mental model of naive users be improved?</h2>
</section>
<section>
<h2 class="r-fit-text">Communication of Technical Concepts</h2>
</section>
<section>
<h2>System Insights for Improved Comprehension</h2>
<div class="side_by_side">
<div style="float:left;">
Relevance Level<br>
image of human and robot (behavior)
</div>
<div style="float:right;">
Architecture Level<br>
image of human and robot (technical insights)
</div>
</div>
<div class="citations">
<span class="citation">Iyad Rahwan et al. “Machine behaviour”. In: Nature 568.7753 (2019),pp. 477-486.</span><br>
<span class="citation">Carsten Schulte and Lea Budde. “A Framework for Computing Education: Hybrid Interaction System: The need for a bigger picture in
computing education”. In: Proceedings of the 18th Koli Calling International Conference on Computing Education Research. 2018, pp. 1-10.</span>
</div>
</section>
<section>
<h2>Ways to Communicate Architecture Insights</h2>
<div class="side_by_side">
<div style="float:left">
Initial Shaping of the Mental Model
<ul>
<li>First contact with a robot</li>
<li>Influence of previous experiences</li>
<li>Priming as Method to form the right mindset</li>
&rArr; <span class="highlight">Instruction</span>
</ul>
</div>
<div style="float:right">
Shaping Mental Models during the Interaction
<ul>
<li>Convey technical concepts</li>
<li>Communication of inner processes</li>
&rArr; <span class="highlight">Process Visualization</span>
</ul>
</div>
</div>
</section>
<section>
<h2>Error Recognition Study</h2>
<ul>
<li>Online user study</li>
<li>Videos of erroneous human-robot interactions</li>
<li>Elaboration on causes of errors</li>
</ul>
</section>
<section>
<h2>Scenario</h2>
<ul>
<li>Interactive Object learning</li>
<li>Utilized Concepts:
<ul>
<li>Object Recognition</li>
<li>Speech Processing</li>
<li>Speech Synthesis</li>
<li>Behavior Control</li>
</ul>
</li>
<!--Add Images for each concept-->
</ul>
</section>
<section>
<h2>Instruction - Goal</h2>
<ul>
<li>Introduce technical concepts <span class="highlight">explicitly</span></li>
<li>Reduce false assumptions</li>
<li>Priming towards a technical mindset</li>
</ul>
</section>
<section>
<h2>Instruction - Realization</h2>
<ul>
<li>Architecture instruction video</li>
<li>Visualization of Robot</li>
<li>Introduce <span class="highlight">Hardware</span> and <span class="highlight">Software</span></li>
</ul>
<!--Add sample frame-->
</section>
<section>
<h2>Process Visualization - Goal</h2>
<ul>
<li>Visualize inner processes <span class="highlight">indirectly</span></li>
<li>Inform but not disturb interaction</li>
<li>Comprehensibility and intuitiveness</li>
</ul>
</section>
<section>
<h2>Process Visualization - Realization</h2>
<div class="r-hstack">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_vis_video.png">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_object_recognition.png">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_robot_chatbox.png">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/kiranshastry_icons.pdf">
</div>
</section>
<section>
<h2>Hypotheses</h2>
<ol>
<li>Hypothesis 1: Providing architectural concepts allows users to gain more knowledge about the functionality of a robot.</li>
<li>Hypothesis 2: Insights into the architecture of a robot increases the ability to recognize and explain occuring error in HRI.</li>
<li>Hypothesis 3: Technical concepts differ in terms of their familiarity and perceivability. These factors influence the user's ability to recognize and understand problems in HRI.</li>
</ol>
</section>
<section>
<h2>Study Design</h2>
<img class="r-stretch" src="media/study1/study1_timeline.svg">
</section>
<section>
<h2>Erroneous Interaction Videos</h2>
<div class="r-hstack">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_interaction.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_no_object.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_failed_speech.svg">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/s1_failed_sm.svg">
<!--Convert images to white; Show one of the videos!-->
</div>
</section>
<section>
<h2>Results - General</h2>
<ul>
<li>Analysis on 122/130 Participants</li>
<li>24 Baselines, 38 Instr, 29 Vis, 31 Instr+Vis</li>
<li>63 women, 56 men, 2 diverse, 1 anonymous</li>
<li>mean age: 34.09 years</li>
</ul>
</section>
<section>
<h2>Results - Hypothesis 1</h2>
<div class="side_by_side">
<ul>
<li></li>
</ul>
<div class="r-hstack">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/hw_num_mentioned.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/hardware_stats_perc.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/sw_num_mentioned.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/software_stats_perc.pdf">
</div>
</div>
</section>
<section>
<h2>Results - Hypothesis 2</h2>
<div class="side_by_side">
<ul>
<li></li>
</ul>
<img class="r-stretch" src="media/study1/Expl_correct.pdf">
</div>
</section>
<section>
<h2>Results - Hypothesis 3</h2>
<div class="side_by_side">
<ul>
<li></li>
</ul>
<div class="r-hstack">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/Expl_correct_grouped_by_problem.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/observation_categories_stats_grouped_by_video.pdf">
</div>
</div>
</section>
<section>
<h2>Results - Subjective Ratings</h2>
<div class="side_by_side">
<ul>
<li></li>
</ul>
<div class="r-hstack">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/Expl_correct_vs_ATI.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/Expl_correct_vs_SUS.pdf">
<img class="fragment fade-in-then-out r-stretch" src="media/study1/Expl_correct_vs_GODSPEED.pdf">
</div>
</div>
</section>
<section>
<h2>Discussion</h2>
<ul>
<li></li>
</ul>
</section>
</section>
<section>
<section>Robot Control Architecture</section>
<section>How do people reason?</section>
<section>Theory Of Mind</section>
<section>Existing Control Architectures</section>
<section>FAMILIAR</section>
</section>
......@@ -102,8 +389,15 @@
</section>
<section>
<section>Discussion of Scientific Quesstions</section>
<section>Outlook</section>
<section>
<h2 class="r-fit-text">Perspectives</h2>
</section>
<section>
<h2>Contributions</h2>
</section>
<section>
<h2>Outlook</h2>
</section>
</section>
</div>
</div>
......
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment