Personalizing Haptics - From Individuals' Sense-Making Schemas to End-User Haptic Tools

Series Editors’ Foreword

Preface

Acknowledgements

Contents

1 Introduction

1.1 Motivation

1.1.1 Leveraging Haptic Utility

1.1.2 Informing Haptic Design and Evaluation

1.2 Situating Our Work

1.2.1 Supporting Personalization

1.2.2 Understanding Common Patterns and Individual Differences

1.2.3 Evaluating at Scale

1.3 Outline of the Chapters

1.3.1 Chapter 2—Linking Emotion Attributes to Engineering Parameters and Individual Differences

1.3.2 Chapter 3—Characterizing Personalization Mechanisms

1.3.3 Chapter 4—Choosing from a Large Library Using Facets

1.3.4 Chapter 5—Deriving Semantics and Interlinkages of Facets

1.3.5 Chapter 6—Crowdsourcing Haptic Data Collection

1.3.6 Chapter 7—Tuning Vibrations with Emotion Controls

1.4 Contributions

1.4.1 Effective Mechanisms for Haptic Personalization

1.4.2 Haptic Facets Encapsulating Common Patterns and Variations in Affect

1.4.3 Methodology for Evaluating Haptic Sensations at a Large Scale

1.4.4 Tools and Datasets

References

2 Linking Emotion Attributes to Engineering Parameters and Individual Differences

2.1 Introduction

2.2 Related Work

2.2.1 Affective Evaluation

2.2.2 Vibrotactile Stimuli

2.2.3 Tactile Tasks

2.2.4 Individual Differences in Tactile Task Performance

2.3 Design of Setup and Assessment Tools

2.3.1 Apparatus

2.3.2 Stimuli Design

2.3.3 Tactile Task Design

2.3.4 Affective Rating Scales Design

2.4 Study

2.4.1 Procedure

2.4.2 Results and Analysis

2.5 Discussion

2.5.1 Dimensionality and Utility of Affective Response

2.5.2 Vibration Parameters

2.5.3 Demographic, NFT Score and Tactile Performance

2.6 Conclusion

References

3 Characterizing Personalization Mechanisms

3.1 Introduction

3.2 Related Work

3.2.1 Haptic Design

3.2.2 Haptic Design Tools

3.2.3 Challenges and Potentials of End-User Personalization

3.3 Conceptualization of Haptic Personalization Tools

3.3.1 Three Personalization Tools

3.3.2 Proposed Tool-Characterization Parameter Space

3.4 Methods

3.5 Results

3.5.1 Comparison of the Tools

3.5.2 Interest in Personalization

3.5.3 Vibrations Designed by Participants

3.6 Discussion

3.6.1 Desirable Characteristics (Q1)

3.6.2 Value and Outcomes (Q2, Q3)

3.6.3 Wizard-of-Oz Approach

3.7 Conclusion

References

4 Choosing From a Large Library Using Facets

4.1 Introduction

4.2 Related Work

4.2.1 Vibrotactile Libraries

4.2.2 Vibrotactile Facets

4.2.3 Inspiration From Visualization and Media Collections

4.3 Library and Facet Construction

4.3.1 Library Population

4.3.2 Visualizing and Managing Diversity During Growth

4.4 VibViz: An Interactive Library Navigation Tool

4.4.1 Requirements

4.4.2 VibViz Interface

4.4.3 Dataset

4.5 User Study

4.6 Results

4.6.1 (Q1) Does VibViz Satisfy Our Design Requirements?

4.6.2 (Q2) How Useful and Interesting Is Each Vibration Facet?

4.7 Discussion

4.7.1 Interface Requirements

4.7.2 Vibrotactile Facets

4.8 Conclusions and Future Work

References

5 Deriving Semantics and Interlinkages of Facets

5.1 Introduction

5.1.1 Facets: Aligning Content Access with Mental Frameworks

5.1.2 Research Questions

5.1.3 Scope

5.2 Related Work

5.2.1 Tools for Vibrotactile Design and Personalization

5.2.2 Knowledge of Perceptual and Qualitative Attributes of Vibrations

5.2.3 Methodology for Evaluating Qualitative Attributes of Vibrations

5.2.4 Instruments for Evaluating Haptic Sensations

5.3 Approach

5.3.1 Rich Source Vibrations

5.3.2 Inclusive and Concise Annotation Instrument, for a Flat Descriptor Set

5.3.3 Scalable and Robust Data Collection Methodology

5.3.4 Data Analysis Methods

5.4 Data Collection and Pre-processing

5.4.1 Stage 1: Annotation by Haptics Experts

5.4.2 Stage 2: Validation of the Dataset by Lay Users

5.4.3 Pre-processing of the Dataset

5.4.4 Definition of Analysis Metrics

5.5 Analysis and Results

5.5.1 [Q1] Facet Substructure: What Are the Underlying Facet Dimensions That Dominate User Reactions to Vibrations?

5.5.2 [Q2] Between-Facet Linkages: How Are Attributes and Dimensions Linked Across Facets?

5.5.3 [Q3] Individual Differences: To What Extent Do People Coincide or Differ in Their Assessment of Vibration Attributes?

5.5.4 Methodology: How Does Staged Data Collection Impact Annotation Quality?

5.6 Discussion

5.6.1 Within-Facet Perceptual Continuity: Scenarios

5.6.2 Facet Dimensions and Linkages

5.6.3 Individuals' Annotation Reliability and Variation

5.6.4 Review of Our Methodology

5.7 Conclusion

References

6 Crowdsourcing Haptic Data Collection

6.1 Introduction

6.2 Related Work

6.2.1 Existing Evaluation Methods for Vibrotactile Effects

6.2.2 Affective Haptics

6.2.3 Mechanical Turk (MTurk)

6.3 Sourcing Reference Vibrations and Qualities

6.3.1 High-Fidelity Reference Library

6.3.2 Affective Properties and Rating Scales

6.4 Proxy Choice and Design

6.4.1 Visualization Design ( and )

6.4.2 Low Fidelity Vibration Design

6.5 Study 1: In-Lab Proxy Vibration Validation (G1)

6.5.1 Comparison Metric: Equivalence Threshold

6.5.2 Proxy Validation (Study 1) Results and Discussion

6.6 Study 2: Deployment Validation with MTurk (G2)

6.6.1 Results

6.7 Discussion

6.7.1 Proxy Modalities Are Viable for Crowdsourcing (G1, G2: Feasibility)

6.7.2 Triangulation (G3: Promising Directions/Proxies)

6.7.3 Animate Visualizations (G3: Promising Directions)

6.7.4 Sound Could Represent Energy (G3: Promising Directions)

6.7.5 Device Dependency and Need for Energy Model for Vibrations (G4: Challenges)

6.7.6 Vibrotactile Affective Ratings Are Generally Noisy (G4: Challenges)

6.7.7 Response and Data Quality for MTurk LofiVib Vibrations (G4: Challenges)

6.7.8 Automatic Translation (G4: Challenges)

6.7.9 Limitations

6.8 Conclusion

References

7 Tuning Vibrations with Emotion Controls

7.1 Introduction

7.1.1 Research Questions, Approach and Contributions

7.2 Related Work

7.2.1 Haptic Design, and Inspirations from Other Domains

7.2.2 Affective Vibration Design

7.3 Starting Points: Use Cases, Initial Vibrations and Linkages

7.3.1 Design and Personalization Use Cases

7.3.2 Choosing Basis Vibrations

7.3.3 Identifying Influential Engineering Parameters

7.4 User Studies

7.4.1 Overview of the User Studies

7.4.2 Methods

7.4.3 Analysis

7.5 Results

7.5.1 Verbal Descriptions for Emotion Attributes

7.5.2 Ratings

7.5.3 RQ1: Impact of Engineering Parameters on Emotion Attributes (Study 1)

7.5.4 RQ2: Evidence of Continuity of the Engineering-Emotion Mappings (Study 2)

7.5.5 RQ3: Impact of Base Vibrations on Emotion Attribute Ratings

7.5.6 RQ4: Orthogonality of Emotion Dimensions

7.6 Discussion

7.6.1 Findings

7.6.2 Automatable Emotion Controls and Study Approach

7.6.3 What Do These Results Enable? Revisiting Our Use Cases

7.6.4 Future Work

7.7 Conclusion

References

8 Conclusion and Future Directions

8.1 Personalization Mechanisms

8.2 Facets as an Underlying Model for Personalization Tools

8.3 Large Scale Evaluation for Theory and Tool Development

8.4 Future Work

8.4.1 Incorporating Personalization in Users' Workflow

8.4.2 Expanding the Mechanisms and the Underlying Model

8.5 Final Remarks

References

9 Supplemental Materials for Chapter5

9.1 List of Tags and Their Disagreement Values

9.2 Tag Removal Summary

9.3 Rating Correlations

9.4 Multidimensional Scaling Graphs on Tag Distances

9.5 Individual Differences in Vibrations

9.6 Between-Facet Tag Linkages

10 Supplemental Materials for Chapter7

10.1 Linkages Between Three Emotion Dimensions and Sensory Attributes of Vibrations

10.2 Implementation of Sensory Parameters

10.3 Qualitative Description of Emotion Dimensions

10.4 Specification of Base Vibrations and Their Derivatives

References