注目の論文一覧

We present the research area of personal dream informatics: studying the self-information systems that support dream engagement and communication between the dreaming self and the wakeful self. Through a survey study of 281 individuals primarily recruited from an online community dedicated to dreaming, we develop a dream-information systems view of dreaming and dream tracking as a type of self-information system. While dream-information systems are characterized by diverse tracking processes, motivations, and outcomes, they are universally constrained by the ephemeral dreamset - the short period of time between waking up and rapid memory loss of dream experiences. By developing a system dynamics model of dreaming we highlight feedback loops that serve as high leverage points for technology designers, and suggest a variety of design considerations for crafting technology that best supports dream recall, dream tracking, and dreamwork for nightmare relief and personal development.

In this paper, we explore how the physically embodied nature of robots can influence learning through non-verbal communication, such as gesturing. We take an embodied cognition perspective to examine student interactions with a NAO robot that uses gestures while reasoning about geometry conjectures. College aged students (N = 30) were randomly assigned to either a dynamic condition, where the robot uses dynamic gestures that represent and manipulate geometric shapes in the conjectures, or control condition, where the robot uses beat gestures that match the rhythm of speech. Students in the dynamic condition: (1) use more gestures when they reason about geometry conjectures, (2) look more at the robot as it speaks, (3) feel the robot is a better study partner and uses effective gestures, but (4) were not more successful in correctly reasoning about geometry conjectures. We discuss implications for socially supported and embodied learning with a physically present robot.

Constructing olfactory interfaces on demand requires significant design proficiency and engineering effort. The absence of powerful and convenient tools that reduced innovation complexity posed obstacles for future research in the area. To address this problem, we proposed O&O, a modular olfactory interface DIY toolkit. The toolkit consists of: (1) a scent generation kit, a set of electronics and accessories that supported three common scent vaporization techniques; (2) a module construction kit, a set of primitive cardboard modules for assembling permutable functional structures; (3) a design manual, a step-by-step design thinking framework that directs the decision-making and prototyping process. We organized a formal workshop with 19 participants and four solo DIY trials to evaluate the capability of the toolkit, the overall user engagement, the creations in both sessions, and the iterative suggestions. Finally, design implications and future opportunities were discussed for further research.

Visual storytelling is a new approach to creative expression based on verbal and figural creativity. The keys to visual storytelling are narrating and drawing over a period of time, which can be beneficial but also demanding on creativity for children. Informed by need-finding investigations, we developed StoryDrawer, a co-creative system that supports visual storytelling for children aged 6–10 years through collaborative drawing between children and artificial intelligence (AI). The system includes a context-based voice agent and two AI-driven collaborative strategies: the real-time transformation of children’s telling into drawings, and the generation of abstract sketches with semantic similarity to existing story content. We conducted a 2 × 2 study with 64 children to evaluate the efficacy of StoryDrawer by varying the two strategies in four conditions. The results suggest that StoryDrawer provoked participants’ creative and elaborate ideas and contributed to their creative outcomes during an engaging visual storytelling experience.

Advances in Natural Language Processing offer techniques to detect the empathy level in texts. To test if individual feedback on certain students’ empathy level in their peer review writing process will help them to write more empathic reviews, we developed ELEA, an adaptive writing support system that provides students with feedback on the cognitive and emotional empathy structures. We compared ELEA to a proven empathy support tool in a peer review setting with 119 students. We found students using ELEA wrote more empathic peer reviews with a higher level of emotional empathy compared to the control group. The high perceived skill learning, the technology acceptance, and the level of enjoyment provide promising results to use such an approach as a feedback application in traditional learning settings. Our results indicate that learning applications based on NLP are able to foster empathic writing skills of students in peer review scenarios.

Textile interfaces enable designers to integrate unobtrusive media and smart home controls into furniture such as sofas. While the technical aspects of such controllers have been the subject of numerous research projects, the physical form factor of these controls has received little attention so far. This work investigates how general design properties, such as overall slider shape, raised vs. recessed sliders, and number and layout of tick marks, affect users' preferences and performance. Our first user study identified a preference for certain design combinations, such as recessed, closed-shaped sliders. Our second user study included performance measurements on variations of the preferred designs from study 1, and took a closer look at tick marks. Tick marks supported orientation better than slider shape. Sliders with at least three tick marks were preferred, and performed well. Non-uniform, equally distributed tick marks reduced the movements users needed to orient themselves on the slider.

Dynamically drawn content (e.g., handwritten text) in learning videos is believed to improve users’ engagement and learning over static powerpoint-based ones. However, evidence from existing literature is inconclusive. With the emergence of Optical Head-Mounted Displays (OHMDs), recent work has shown that video learning can be adapted for on-the-go scenarios. To better understand the role of dynamic drawing, we decoupled dynamically drawn text into two factors (font style and motion of appearance) and studied their impact on learning performance under two usage scenarios (while seated with desktop and walking with OHMD). We found that although letter-traced text was more engaging for some users, most preferred learning with typeface text that displayed the entire word at once and achieved better recall (46.7% higher), regardless of the usage scenarios. Insights learned from the studies can better inform designers on how to present text in videos for ubiquitous access.

Despite an abundance of carefully-crafted tutorials, trial-and-error remains many people’s preferred way to learn complex software. Yet, approaches to facilitate trial-and-error (such as tooltips) have evolved very little since the 1980s. While existing mechanisms work well for simple software, they scale poorly to large feature-rich applications. In this paper, we explore new techniques to support trial-and-error in complex applications. We identify key benefits and challenges of trial-and-error, and introduce a framework with a conceptual model and design space. Using this framework, we developed three techniques: ToolTrack to keep track of trial-and-error progress; ToolTrip to go beyond trial-and-error of single commands by highlighting related commands that are frequently used together; and ToolTaste to quickly and safely try commands. We demonstrate how these techniques facilitate trial-and-error, as illustrated through a proof-of-concept implementation in the CAD software Fusion 360. We conclude by discussing possible scenarios and outline directions for future research on trial-and-error.

Guided by the 12 principles of animation, stylization is a core 2D animation feature but has been utilized mainly by experienced animators. Although there are tools for stylizing 2D animations, creating stylized 3D animations remains a challenging problem due to the additional spatial dimension and the need for responsive actions like contact and collision. We propose a system that helps users create stylized casual 3D animations. A layered authoring interface is employed to balance between ease of use and expressiveness. Our surface level UI is a timeline sequencer that lets users add preset stylization effects such as squash and stretch and follow through to plain motions. Users can adjust spatial and temporal parameters to fine-tune these stylizations. These edits are propagated to our node-graph-based second level UI, in which the users can create custom stylizations after they are comfortable with the surface level UI. Our system also enables the stylization of interactions among multiple objects like force, energy, and collision. A pilot user study has shown that our fluid layered UI design allows for both ease of use and expressiveness better than existing tools.

Electrotactile stimulation is a novel form of haptic feedback. There is little work investigating its basic design parameters and how they create effective tactile cues. This paper describes two experiments that extend our knowledge of two key parameters. The first investigated the combination of pulse width and amplitude Intensity on sensations of urgency, annoyance, valence and arousal. Results showed significant effects: increasing Intensity caused higher ratings of urgency, annoyance and arousal but reduced valence. We established clear levels for differentiating each sensation. A second study then investigated Intensity and Pulse Frequency to find out how many distinguishable levels could be perceived. Results showed that both Intensity and Pulse Frequency significantly affected perception, with four distinguishable levels of Intensity and two of Pulse Frequency. These results add significant new knowledge about the parameter space of electrotactile cue design and help designers select suitable properties to use when creating electrotactile cues.

How-to videos are often shot using camera angles that may not be optimal for learning motor tasks, with a prevalent use of third-person perspective. We present \textit{immersivePOV}, an approach to film how-to videos from an immersive first-person perspective using a head-mounted 360° action camera. immersivePOV how-to videos can be viewed in a Virtual Reality headset, giving the viewer an eye-level viewpoint with three Degrees of Freedom. We evaluated our approach with two everyday motor tasks against a baseline first-person perspective and a third-person perspective. In a between-subjects study, participants were assigned to watch the task videos and then replicate the tasks. Results suggest that immersivePOV reduced perceived cognitive load and facilitated task learning. We discuss how immersivePOV can also streamline the video production process for content creators. Altogether, we conclude that immersivePOV is an effective approach to film how-to videos for learners and content creators alike.

Most video-based learning content is designed for desktops without considering mobile environments. We (1) investigate the gap between mobile learners’ challenges and video engineers’ considerations using mixed methods and (2) provide design guidelines for creating mobile-friendly MOOC videos. To uncover learners’ challenges, we conducted a survey (n=134) and interviews (n=21), and evaluated the mobile adequacy of current MOOCs by analyzing 41,722 video frames from 101 video lectures. Interview results revealed low readability and situationally-induced impairments as major challenges. The content analysis showed a low guideline compliance rate for key design factors. We then interviewed 11 video production engineers to investigate design factors they mainly consider. The engineers mainly focus on the size and amount of content while lacking consideration for color, complex images, and situationally-induced impairments. Finally, we present and validate guidelines for designing mobile-friendly MOOCs, such as providing adaptive and customizable visual design and context-aware accessibility support.

Optical see-through Head-Mounted Displays (OST HMDs, OHMDs) are known to facilitate situational awareness while accessing secondary information. However, information displayed on OHMDs can cause attention shifts, which distract users from natural social interactions. We hypothesize that information displayed in paracentral and near-peripheral vision can be better perceived while the user is maintaining eye contact during face-to-face conversations. Leveraging this idea, we designed a circular progress bar to provide progress updates in paracentral and near-peripheral vision. We compared it with textual and linear progress bars under two conversation settings: a simulated one with a digital conversation partner and a realistic one with a real partner. Results show that a circular progress bar can effectively reduce notification distractions without losing eye contact and is more preferred by users. Our findings highlight the potential of utilizing the paracentral and near-peripheral vision for secondary information presentation on OHMDs.

With the rapid development of creativity support tools, creative practitioners (e.g., designers, artists, architects) have to constantly explore and adopt new tools into their practice. While HCI research has focused on developing novel creativity support tools, little is known about creative practitioner's values when exploring and adopting these tools. We collect and analyze 23 videos, 13 interviews, and 105 survey responses of creative practitioners reflecting on their values to derive a value framework. We find that practitioners value the tools' functionality, integration into their current workflow, performance, user interface and experience, learning support, costs and emotional connection, in that order. They largely discover tools through personal recommendations. To help unify and encourage reflection from the wider community of CST stakeholders (e.g., systems creators, researchers, marketers, educators), we situate the framework within existing research on systems, creativity support tools and technology adoption.

FabricatINK explores the personal fabrication of irregularly-shaped low-power displays using electronic ink (E ink). E ink is a programmable bicolour material used in traditional form-factors such as E readers. It has potential for more versatile use within the scope of personal fabrication of custom-shaped displays, and it has the promise to be the pre-eminent material choice for this purpose. We appraise technical literature to identify properties of E ink, suited to fabrication. We identify a key roadblock, universal access to E ink as a material, and we deliver a method to circumvent this by upcycling broken electronics. We subsequently present a novel fabrication method for irregularly-shaped E ink displays. We demonstrate our fabrication process and E ink's versatility through ten prototypes showing different applications and use cases. By addressing E ink as a material for display fabrication, we uncover the potential for users to create custom-shaped truly bistable displays.

We investigate performance characteristics when switching between four pointing methods: absolute touch, absolute pen, relative mouse, and relative trackpad. The established "subtraction method" protocol used in mode-switching studies is extended to test pairs of methods and accommodate switch direction, multiple baselines, and controlling relative cursor position. A first experiment examines method switching on and around the horizontal surface of a tablet. Results find switching between pen and touch is fastest, and switching between relative and absolute methods incurs additional time penalty. A second experiment expands the investigation to an emerging foldable all-screen laptop form factor where switching also occurs on an angled surface and along a smoothly curved hinge. Results find switching between trackpad and touch is fastest, with all switching times generally higher. Our work contributes missing empirical evidence for switching performance using modern input methods, and our results can inform interaction design for current and emerging device form factors.

Prior work has demonstrated augmented reality's benefits to education, but current tools are difficult to integrate with traditional instructional methods. We present Paper Trail, an immersive authoring system designed to explore how to enable instructors to create AR educational experiences, leaving paper at the core of the interaction and enhancing it with various forms of digital media, animations for dynamic illustrations, and clipping masks to guide learning. To inform the system design, we developed five scenarios exploring the benefits that hand-held and head-worn AR can bring to STEM instruction and developed a design space of AR interactions enhancing paper based on these scenarios and prior work. Using the example of an AR physics handout, we assessed the system's potential with PhD-level instructors and its usability with XR design experts. In an elicitation study with high-school teachers, we study how Paper Trail could be used and extended to enable flexible use cases across various domains. We discuss benefits of immersive paper for supporting diverse student needs and challenges for making effective use of AR for learning.

Making time estimates, such as how long a given task might take, frequently leads to inaccurate predictions because of an optimistic bias. Previous attempts to alleviate this bias, including decomposing the task into smaller components and listing potential surprises, have not shown any major improvement. This article builds on the premise that these procedures may have failed because they involve compound probabilities and mixture distributions which are difficult to compute in one's head. We hypothesize that predictive visualizations of such distributions would facilitate the estimation of task durations. We conducted a crowdsourced study in which 145 participants provided different estimates of overall and sub-task durations and we used these to generate predictive visualizations of the resulting mixture distributions. We compared participants' initial estimates with their updated ones and found compelling evidence that predictive visualizations encourage less optimistic estimates.

Microtransactions have become a major monetisation model in digital games, shaping their design, impacting their player experience, and raising ethical concerns. Research in this area has chiefly focused on loot boxes. This begs the question whether other microtransactions might actually be more relevant and problematic for players. We therefore conducted a content analysis of negative player reviews (n=801) of top-grossing mobile and desktop games to determine which problematic microtransactions are most prevalent and salient for players. We found that problematic microtransactions with mobile games featuring more frequent and different techniques compared to desktop games. Across both, players minded issues related to fairness, transparency, and degraded user experience, supporting prior theoretical work, and importantly take issue with monetisation-driven design as such. We identify future research needs on why microtransactions in particular spark this critique, and which player communities it may be more or less representative of.

Fluid fiber, with fluid flowing in a tube, is an attractive material to flexibly and dynamically display digital information. While it has recently attracted attention from the HCI field, there is currently little knowledge about this material, limited to controlling the position of the droplets to present representational information like letters and numbers. To develop a broader and deepened understanding of this material and its potential for display design, we conducted a study based on a design workshop where art and design practitioners engaged in creation practice with a toolkit we designed and developed. The toolkit includes hardware components for controlling bubbles and droplets and a GUI design tool for arranging the fluid layout. Our research reveals the structural and expressive affordance of such a fluid fiber for displaying information, highlighting the unique value of fluidity as an intuitive form to express life, emotion, movement and changes.

Today spectating and streaming virtual reality (VR) activities typically involves spectators viewing a 2D stream of the VR user’s view. Streaming 2D videos of the game play is popular and well-supported by platforms such as Twitch. However, the generic streaming of full 3D representations is less explored. Thus, while the VR player’s experience may be fully immersive, spectators are limited to 2D videos. This asymmetry lessens the overall experience for spectators, who themselves may be eager to spectate in VR. DreamStream puts viewers in the virtual environment of the VR application, allowing them to look “over the shoulder” of the VR player. Spectators can view streamed VR content immersively in 3D, independently explore the VR scene beyond what the VR player sees and ultimately cohabit the virtual environment alongside the VR player. For the VR player, DreamStream provides a spatial awareness of all their spectators. DreamStream retrofits and works with existing VR applications. We discuss the design and implementation of DreamStream, and carry out three qualitative informal evaluations. These evaluations shed light on the strengths and weakness of using DreamStream for the purpose of interactive spectating. Our participants found that DreamStream’s VR viewer interface offered increased immersion, and made it easier to communicate and interact with the VR player.

Speech is inappropriate in many situations, limiting when voice control can be used. Most unvoiced speech text entry systems can not be used while on-the-go due to movement artifacts. Using a dental retainer with capacitive touch sensors, SilentSpeller tracks tongue movement, enabling users to type by spelling words without voicing. SilentSpeller achieves an average 97% character accuracy in offline isolated word testing on a 1164-word dictionary. Walking has little effect on accuracy; average offline character accuracy was roughly equivalent on 107 phrases entered while walking (97.5%) or seated (96.5%). To demonstrate extensibility, the system was tested on 100 unseen words, leading to an average 94% accuracy. Live text entry speeds for seven participants averaged 37 words per minute at 87% accuracy. Comparing silent spelling to current practice suggests that SilentSpeller may be a viable alternative for silent mobile text entry.

Data exploration systems have become popular tools with which data analysts and others can explore raw data and organize their observations. However, users of such systems who are unfamiliar with their datasets face several challenges when trying to extract data events of interest to them. Those challenges include progressively discovering informative charts, organizing them into a logical order to depict a meaningful fact, and arranging one or more facts to illustrate a data event. To alleviate them, we propose VisGuide - a data exploration system that generates personalized recommendations to aid users’ discovery of data events in breadth and depth by incrementally learning their data exploration preferences and recommending meaningful charts tailored to them. As well as user preferences, VisGuide’s recommendations simultaneously consider sequence organization and chart presentation. We conducted two user studies to evaluate 1) the usability of VisGuide and 2) user satisfaction with its recommendation system. The results of those studies indicate that VisGuide can effectively help users create coherent and user-oriented visualization trees that represent meaningful data events.

What do pedestrian crossings, ATMs, elevators and ticket machines have in common? These are just a few of the ubiquitous yet essential elements of public-space infrastructure that rely on physical buttons or touchscreens; common interactions that, until recently, were considered perfectly safe to perform. This work investigates how we might integrate touchless technologies into public-space infrastructure in order to minimise physical interaction with shared devices in light of the ongoing COVID-19 pandemic. Drawing on an ethnographic exploration into how public utilities are being used, adapted or avoided, we developed and evaluated a suite of technology probes that can be either retro tted into, or replace, these services. In-situ community deployments of our probes demonstrate strong uptake and provide insight into how hands-free technologies can be adapted and utilised for the public domain; and, in turn, used to inform the future of walk-up-and use public technologies.

We investigate instant-messaging (IM) users’ sense-making and practices around read-receipts: a feature of IM apps for supporting the awareness of turn-taking, i.e., whether a message recipient has read a message. Using a grounded-theory approach, we highlight the importance of five contextual factors – situational, relational, interactional, conversational, and personal – that shape the variety of IM users’ sense-making about read-receipts and strategies for utilizing them in different settings. This approach yields a 21-part typology comprising five types of senders’ speculation about why their messages with read-receipts have not been answered; eight types of recipients’ causes/reasons behind such non-response; and four types of senders’ and recipients’ subsequent strategies, respectively. Mismatches between senders’ speculations about un-responded-to read-receipted messages (URRMs) and recipients’ self-reported explanations are also discussed as sources of communicative friction. The findings reveal that, beyond indicating turn-taking, read-receipts have been leveraged as a strategic tool for various purposes in interpersonal relations.

People with limited mobility often use multiple devices when interacting with computing systems, but little is known about the impact these multi-modal configurations have on daily computing use. A deeper understanding of the practices, preferences, obstacles, and workarounds associated with accessible multi-modal input can uncover opportunities to create more accessible computer applications and hardware. We explored how people with limited mobility use multi-modality through a three-part investigation grounded in the context of video games. First, we surveyed 43 people to learn about their preferred devices and configurations. Next, we conducted semi-structured interviews with 14 participants to understand their experiences and challenges with using, configuring, and discovering input setups. Lastly, we performed a systematic review of 74 YouTube videos to illustrate and categorize input setups and adaptations in-situ. We conclude with a discussion on how our findings can inform future accessibility research for current and emerging computing technologies.

We present STRAIDE, a string-actuated interactive display environment that allows to explore the promising potential of shape-changing interfaces for casual visualizations. At the core, we envision a platform that spatially levitates elements to create dynamic visual shapes in space. We conceptualize this type of tangible mid-air display and discuss its multifaceted design dimensions. Through a design exploration, we realize a physical research platform with adjustable parameters and modular components. For conveniently designing and implementing novel applications, we provide developer tools ranging from graphical emulators to in-situ augmented reality representations. To demonstrate STRAIDE's reconfigurability, we further introduce three representative physical setups as a basis for situated applications including ambient notifications, personal smart home controls, and entertainment. They serve as a technical validation, lay the foundations for a discussion with developers that provided valuable insights, and encourage ideas for future usage of this type of appealing interactive installation.

In recognition of food’s significant experiential pleasures, culinary practitioners and designers are increasingly exploring novel combinations of computing technologies and food. However, despite much creative endeavors, proposals and prototypes have so far largely maintained a traditional divide, treating food and technology as separate entities. In contrast, we present a “Research through Design” exploration of the notion of food as computational artifact: wherein food itself is the material of computation. We describe the Logic Bonbon, a dessert that can hydrodynamically regulate its flavor via a fluidic logic system. Through a study of experiencing the Logic Bonbon and reflection on our design practice, we offer a provisional account of how food as computational artifact can mediate new interactions through a novel approach to food-computation integration, that promotes an enriched future of Human-Food Interaction.

Telephone-driven community forums have been a widely proposed solution to address the unreliable internet connectivity and large geographic scope that characterizes many international NGO contexts. Primarily, these applications support asynchronous activities, such as information portals or forums to access rural journalism, but opportunities for real-time experience sharing remain largely under-explored. In this paper, we explore the potential of such forums to support remote mentoring of NGO volunteers, a practice that requires synchronous, dialogical formats for experience sharing and peer discussion. We engaged 28 participants from a rural Indian NGO in the design of peer-mentoring sessions that leverage synchronous audio discussions, using the structure and format of traditional talk-show radio as a starting point. The participants favored an entertaining approach to mentoring and discussed the logistics required to achieve this within their resource constraints. We conclude with design implications for designing media-driven community engagement platforms and the ethical challenges around protecting marginalized community interests.

Rendering instant and intense impact feedback on users’ hands, limbs and head to enhance realism in virtual reality (VR) has been proposed in previous works, but impact on the body is still less discussed. With the body’s large surface area to utilize, numerous impact patterns can be rendered in versatile VR applications,e.g., being shot, blasted, punched or slashed on body in VR games. Herein we propose ImpactVest to render spatio-temporal multilevel impact force feedback on body. By independently controlling nine impactors in a 3×3 layout using elastic force, impact is generated at different levels, positions and time sequences for versatile spatial and temporal combinations. We conducted a just-noticeable difference (JND) study to understand users’ impact level distinguishability on the body. A time interval threshold study was then performed to ascertain what time interval thresholds between two impact stimuli should be used to distinguish from simultaneous impact, a continuous impact stroke and two discrete impact stimuli. Based on the results, we conducted a VR experience study to verify that impact feedback from ImpactVest enhances VR realism.

Visuo-haptic illusions are a method to expand proxy-based interactions in VR by introducing unnoticeable discrepancies between the virtual and real-world. Yet, how different design variables affect the illusions with proxies is still unclear. To unpack a subset of variables, we conducted two user studies with 48 participants to explore the impact of (1) different grasping types and movement trajectories, and (2) different grasping types and object masses on the discrepancy which may be introduced. Our Bayes analysis suggests that grasping types and object masses (≤ 500 g) did not noticeably affect the discrepancy, but for movement trajectory, results were inconclusive. Further, we identified a significant difference between un-/restricted movement trajectories. Our data shows considerable differences in participants’ proprioceptive accuracy, which seem to correlate with their prior VR experience. Finally, we illustrate the impact of our key findings on the visuo-haptic illusion design process by showcasing a new design workflow.

The effective width method of Fitts' law can normalize speed-accuracy biases in 1D target pointing tasks. However, in graphical user interfaces, more meaningful target shapes are rectangular. To empirically determine the best way to normalize the subjective biases, we ran remote and crowdsourced user experiments with three speed-accuracy instructions. We propose to normalize the speed-accuracy biases by applying the effective sizes to existing Fitts' law formulations including width W and height H. We call this target-size adjustment the bivariate effective width method. We found that, overall, Accot and Zhai's weighted Euclidean model using the effective width and height independently showed the best fit to the data in which the three instruction conditions were mixed (i.e., the time data measured in all instructions were analyzed with a single regression expression). Our approach enables researchers to fairly compare two or more conditions (e.g., devices, input techniques, user groups) with the normalized throughputs.

Data visualizations are created and shared on the web at an unprecedented speed, raising new needs and questions for processing and analyzing visualizations after they have been generated and digitized. However, existing formalisms focus on operating on a single visualization instead of multiple visualizations, making it challenging to perform analysis tasks such as sorting and clustering visualizations. Through a systematic analysis of previous work, we abstract visualization-related tasks into mathematical operators such as union and propose a design space of visualization operations. We realize the design by developing ComputableViz, a library that supports operations on multiple visualization specifications. To demonstrate its usefulness and extensibility, we present multiple usage scenarios concerning processing and analyzing visualization, such as generating visualization embeddings and automatically making visualizations accessible. We conclude by discussing research opportunities and challenges for managing and exploiting the massive visualizations on the web.

The diminutive size of wrist wearables has prompted the design of many novel input techniques to increase expressivity. Finger identification, or assigning different functionality to different fingers, has been frequently proposed. However, while the value of the technique seems clear, its implementation remains challenging, often relying on external devices (e.g., worn magnets) or explicit instructions. Addressing these limitations, this paper explores a novel approach to natural and unencumbered finger identification on an unmodified smartwatch: sonar. To do this, we adapt an existing finger tracking smartphone sonar implementation---rather than extract finger motion, we process raw sonar fingerprints representing the complete sonar scene recorded during a touch. We capture data from 16 participants operating a smartwatch and use their sonar fingerprints to train a deep learning recognizer that identifies taps by the thumb, index, and middle fingers with an accuracy of up to 93.7%, sufficient to support meaningful application development.

Now more than ever, people are using online platforms to communicate. Twitch, the foremost platform for live game streaming, offers many communication modalities. However, the platform lacks representation of social cues and signals of the audience experience, which are innately present in live events. To address this, we present a technology probe that captures the audience energy and response in a game streaming context. We designed a game and integrated a custom-communication modality—Commons Sense—in which the audience members' heart rates are sensed via webcam, averaged, and fed into a video game to affect sound, lighting, and difficulty. We conducted an `in-the-wild' evaluation with four Twitch streamers and their audience members (N=55) to understand how these groups interacted through Commons Sense. Audience members and streamers indicated high levels of enjoyment and engagement with Commons Sense, suggesting the potential of physiological interaction as a beneficial communication tool in live streaming.

Existing conversational approaches for Internet of Things (IoT) service mashup do not support modification because of the usability challenge, although it is common for users to modify the service mashups in IoT environments. To support the modification of IoT service mashups through conversational interfaces in a usable manner, we propose the conversational mashup modification agent (CoMMA). Users can modify IoT service mashups using CoMMA through natural language conversations. CoMMA has a two-step mashup modification interaction, an implicature-based localization step, and a modification step with a disambiguation strategy. The localization step allows users to easily search for a mashup by vocalizing their expressions in the environment. The modification step supports users to modify mashups by speaking simple modification commands. We conducted a user study and the results show that CoMMA is as effective as visual approaches in terms of task completion time and perceived task workload for modifying IoT service mashups.

From creating input devices to rendering tangible information, the field of HCI is interested in using kinematic mechanisms to create human-computer interfaces. Yet, due to fabrication and design challenges, it is often difficult to create kinematic devices that are compact and have multiple reconfigurable motional degrees of freedom (DOFs) depending on the interaction scenarios. In this work, we combine compliant mechanisms (CMs) with tensioning cables to create dynamically reconfigurable kinematic mechanisms. The devices’ kinematics (DOFs) is enabled and determined by the layout of bendable rods. The additional cables function as on-demand motion constraints that can dynamically lock or unlock the mechanism’s DOFs as they are tightened or loosened. We provide algorithms and a design tool prototype to help users design such kinematic devices. We also demonstrate various HCI use cases including a kinematic haptic display, a haptic proxy, and a multimodal input device.

Soft actuators with integrated sensing have shown utility in a variety of applications such as assistive wearables, robotics, and interactive input devices. Despite their promise, these actuators can be difficult to both design and fabricate. As a solution, we present a workflow for computationally designing and digitally fabricating soft pneumatic actuators \emph{via} a machine knitting process. Machine knitting is attractive as a fabrication process because it is fast, digital (programmable), and provides access to a rich material library of functional yarns for specified mechanical behavior and integrated sensing. Our method uses elastic stitches to construct non-homogeneous knitting structures, which program the bending of actuators when inflated. Our method also integrates pressure and swept frequency capacitive sensing structures using conductive yarns. The entire knitted structure is fabricated automatically in a single machine run. We further provide a computational design interface for the user to interactively preview actuators’ quasi-static shape when authoring elastic stitches. Our sensing-integrated actuators are cost-effective, easy to design, robust to large actuation, and require minimal manual post-processing. We demonstrate five use-cases of our actuators in relevant application settings.

In this paper, we present a novel screen system employing polarizers that allow switching of the projection surface to the front, rear, or both sides using only two projectors on one side. In this system, we propose a method that employs two projectors equipped with polarizers and a multi-layered screen comprising an anti-reflective plate, transparent screen, and wire grid polarizer. The multi-layered screen changes whether the projected image is shown on the front or rear side of the screen depending on the polarization direction of the incident light. Hence, the proposed method can project images on the front, rear, or both sides of the screen by projecting images from either or both projectors using polarizers. In addition, the proposed method can be easily deployed by simply attaching multiple optical films. We implement a prototype and confirm that the proposed method can selectively switch the projection surface.

This paper presents FlexHaptics, a design method for creating custom haptic input interfaces. Our approach leverages planar compliant structures whose force-deformation relationship can be altered by adjusting the geometries. Embedded with such structures, a FlexHaptics module exerts a fine-tunable haptic effect (i.e., resistance, detent, or bounce) along a movement path (i.e., linear, rotary, or ortho-planar). These modules can work separately or combine into an interface with complex movement paths and haptic effects. To enable the parametric design of FlexHaptic modules, we provide a design editor that converts user-specified haptic properties into underlying mechanical structures of haptic modules. We validate our approach and demonstrate the potential of FlexHaptic modules through six application examples, including a slider control for a painting application and a piano keyboard interface on touchscreens, a tactile low vision timer, VR game controllers, and a compound input device of a joystick and a two-step button.

Online learners are hugely diverse with varying prior knowledge, but most instructional videos online are created to be one-size-fits-all. Thus, learners may struggle to understand the content by only watching the videos. Providing scaffolding prompts can help learners overcome these struggles through questions and hints that relate different concepts in the videos and elicit meaningful learning. However, serving diverse learners would require a spectrum of scaffolding prompts, which incurs high authoring effort. In this work, we introduce Promptiverse, an approach for generating diverse, multi-turn scaffolding prompts at scale, powered by numerous traversal paths over knowledge graphs. To facilitate the construction of the knowledge graphs, we propose a hybrid human-AI annotation tool, Grannotate. In our study (N=24), participants produced 40 times more on-par quality prompts with higher diversity, through Promptiverse and Grannotate, compared to hand-designed prompts. Promptiverse presents a model for creating diverse and adaptive learning experiences online.

By sensing how a user is holding a smartphone, adaptive user interfaces are possible such as those that automatically switch the displayed content and position of graphical user interface (GUI) components following how the phone is being held. We propose ReflecTouch, a novel method for detecting how a smartphone is being held by capturing images of the smartphone screen reflected on the cornea with a built-in front camera. In these images, the areas where the user places their fingers on the screen appear as shadows, which makes it possible to estimate the grasp posture. Since most smartphones have a front camera, this method can be used regardless of the device model; in addition, no additional sensor or hardware is required. We conducted data collection experiments to verify the classification accuracy of the proposed method for six different grasp postures, and the accuracy was 85%.

A significant drawback of text passwords for end-user authentication is password reuse. We propose a novel approach to detect password reuse by leveraging gaze as well as typing behavior and study its accuracy. We collected gaze and typing behavior from 49 users while creating accounts for 1) a webmail client and 2) a news website. While most participants came up with a new password, 32% reported having reused an old password when setting up their accounts. We then compared different ML models to detect password reuse from the collected data. Our models achieve an accuracy of up to 87.7% in detecting password reuse from gaze, 75.8% accuracy from typing, and 88.75% when considering both types of behavior. We demonstrate that \revised{using gaze, password} reuse can already be detected during the registration process, before users entered their password. Our work paves the road for developing novel interventions to prevent password reuse.

In voice-based interfaces, non-verbal features represent a simple and underutilized design space for hands-free, language-agnostic interactions. We evaluate the performance of three fundamental types of voice-based musical interactions: pitch, interval, and melody. These interactions involve singing or humming a sequence of one or more notes. A 21-person study evaluates the feasibility and enjoyability of these interactions. The top performing participants were able to perform all interactions reasonably quickly (<5s) with average error rates between 1.3% and 8.6% after training. Others improved with training but still had error rates as high as 46% for pitch and melody interactions. The majority of participants found all tasks enjoyable. Using these results, we propose design considerations for using singing interactions as well as potential use cases for both standard computers and augmented reality glasses.

Haptic Feedback is essential for lifelike Virtual Reality (VR) experiences. To provide a wide range of matching sensations of being touched or stroked, current approaches typically need large numbers of different physical textures. However, even advanced devices can only accommodate a limited number of textures to remain wearable. Therefore, a better understanding is necessary of how expectations elicited by different visualizations affect haptic perception, to achieve a balance between physical constraints and great variety of matching physical textures. In this work, we conducted an experiment (N=31) assessing how the perception of roughness is affected within VR. We designed a prototype for arm stroking and compared the effects of different visualizations on the perception of physical textures with distinct roughnesses. Additionally, we used the visualizations' real-world materials, no-haptics and vibrotactile feedback as baselines. As one result, we found that two levels of roughness can be sufficient to convey a realistic illusion.

The confusion matrix, a ubiquitous visualization for helping people evaluate machine learning models, is a tabular layout that compares predicted class labels against actual class labels over all data instances. We conduct formative research with machine learning practitioners at Apple and find that conventional confusion matrices do not support more complex data-structures found in modern-day applications, such as hierarchical and multi-output labels. To express such variations of confusion matrices, we design an algebra that models confusion matrices as probability distributions. Based on this algebra, we develop Neo, a visual analytics system that enables practitioners to flexibly author and interact with hierarchical and multi-output confusion matrices, visualize derived metrics, renormalize confusions, and share matrix specifications. Finally, we demonstrate Neo's utility with three model evaluation scenarios that help people better understand model performance and reveal hidden confusions.

Social Media (SM) has shown that we adapt our communication and disclosure behaviors to available technological opportunities. Head-mounted Augmented Reality (AR) will soon allow to effortlessly display the information we disclosed not isolated from our physical presence (e.g., on a smartphone) but visually attached to the human body. In this work, we explore how the medium (AR vs. Smartphone), our role (being augmented vs. augmenting), and characteristics of information types (e.g., level of intimacy, self-disclosed vs. non-self-disclosed) impact the users' comfort when displaying personal information. Conducting an online survey (N=148), we found that AR technology and being augmented negatively impacted this comfort. Additionally, we report that AR mitigated the effects of information characteristics compared to those they had on smartphones. In light of our results, we discuss that information augmentation should be built on consent and openness, focusing more on the comfort of the augmented rather than the technological possibilities.

Previous work hypothesized that for Virtual Reality (VR) and Augmented Reality (AR) displays a mismatch between disparities and optical focus cues, known as the vergence and accommodation conflict (VAC), affects depth perception and thus limits user performance in 3D selection tasks within arm's reach (peri-personal space). To investigate this question, we built a multifocal stereo display, which can eliminate the influence of the VAC for pointing within the investigated distances. In a user study, participants performed a virtual hand 3D selection task with targets arranged laterally or along the line of sight, with and without a change in visual depth, in display conditions with and without the VAC. Our results show that the VAC influences 3D selection performance in common VR and AR stereo displays and that multifocal displays have a positive effect on 3D selection performance with a virtual hand.

Sharing annotations encourages feedback, discussion, and knowledge passing among readers and can be beneficial for personal and public use. Prior augmented reality (AR) systems have expanded these benefits to both digital and printed documents. However, despite smartphone AR now being widely available, there is a lack of research about how to use AR effectively for interactive document annotation. We propose Dually Noted, a smartphone-based AR annotation system that recognizes the layout of structural elements in a printed document for real-time authoring and viewing of annotations. We conducted experience prototyping with eight users to elicit potential benefits and challenges within smartphone AR, and this informed the resulting Dually Noted system and annotation interactions with the document elements. AR annotation is often unwieldy, but during a 12-user empirical study our novel structural understanding component allows Dually Noted to improve precise highlighting and annotation interaction accuracy by 13%, increase interaction speed by 42%, and significantly lower cognitive load over a baseline method without document layout understanding. Qualitatively, participants commented that Dually Noted was a swift and portable annotation experience. Overall, our research provides new methods and insights for how to improve AR annotations for physical documents.

Imagine in the future people comfortably wear augmented reality (AR) displays all day, how do we design interfaces that adapt to the contextual changes as people move around? In current operating systems, the majority of AR content defaults to staying at a fixed location until being manually moved by the users. However, this approach puts the burden of user interface (UI) transition solely on users. In this paper, we first ran a bodystorming design workshop to capture the limitations of existing manual UI transition approaches in spatially diverse tasks. Then we addressed these limitations by designing and evaluating three UI transition mechanisms with different levels of automation and controllability (low-effort manual, semi-automated, fully-automated). Furthermore, we simulated imperfect contextual awareness by introducing prediction errors with different costs to correct them. Our results provide valuable lessons about the trade-offs between UI automation levels, controllability, user agency, and the impact of prediction errors.