Papers & Notes

Expand Collapse

We present a survey on touch-based gestures recognition techniques and frameworks. We propose an extended set of requirements such techniques and frameworks should meet in order to provide better support to multi-touch applications. Gestures may be described formally or by user-provided examples. Frameworks integrate the multi-touch sensors drivers to the application and may use different gesture recognizers.

To mitigate multi-device interaction problems (such as lack of control, intelligibility and context), ActivitySpace provides an activity-centric configuration space that enables users to integrate and work across several devices by using the space between the devices. We report on a study with 9 participants that shows that ActivitySpace helps users to easily manage devices and their allocated resources while exposing a number of usage patterns.

ACTO is a customizable, reusable actuated tangible user interface object. Its modular design allows quick adaptations for different tabletop scenarios, making otherwise integral parts like the actuation mechanism or the physical configuration interchangeable. This qualifies it as an ideal research and education platform for tangible user interfaces.

This paper provides a characterization of input force for common touch gestures on mobile devices. This characterization is intended to aid the design of interactive devices that integrate force-input with common touch gestures in different conditions.

Tangibles on interactive surfaces enable users to physically manipulate digital content by placing, manipulating, or removing a tangible object. However, the information whether and how a user grasps these objects has not been mapped out for tangibles on interactive surfaces so far. Based on Buxton’s Three-State Model for graphical input, we present an interaction model that describes input on tangibles that are aware of the user’s grasp.

BullsEye improves the precision of optical fiducial tracking on tangible tabletops to sub-pixel accuracy down to a tenth of a pixel. Techniques include a fiducial design for GPU based tracking, calibration of light that allows for computation on a greyscale image, and an automated technique for optical distortion compensation.

We conducted a survey of 1515 electronic push buttons in everyday home environments. We report a characterisation that describes the features of the data set and discusses important button properties that we expect will inform the design of future physically-dynamic devices and surfaces.

This paper presents a remote collaboration system that enables a remote expert to guide a local worker during a technical support task. The system was designed for industrial settings, which introduced specific requirements for the design. To validate the applicability of the system in a real setting, we conducted a user study with operators and field workers that used the system in their working environment in a water treatment plant.

We present a novel human tracking and activity analysis approach using a top-down 3D camera. Our hierarchical tracking models local and global affinities, scene constraints and motion patterns to track people in space, and a novel salience occupancy pattern (SOP) is used for action recognition. We validate the tracker using two different interactive surface applications: Proximity Table and My Mother’s Kitchen Exhibit.

In collaboration with a military and emergency response simulation software company, we developed ePlan Multi-Surface, a multi-surface environment for communication and collaboration for emergency response planning exercises. We describe the domain, how it informed our prototype, and insights on collaboration, interactions and information dissemination in multi-surface environments for emergency operations centres.

This work investigates narrative hand gestures used on digital surfaces while interacting with virtual objects to narrate a story. We found that narrative gestures are fundamentally different than traditional gestures used to interact with on-screen objects. Narrative gestures are used to communicate visual meaning to an audience.

In this paper, we evaluate two approaches to communicate three different touch gestures (tap, drag and scale up) to pre-kindergarten users. Our results show, firstly, that it is possible to effectively communicate them using visual cues and, secondly, that an animated semiotic approach is better than an iconic one.

We introduce HuddleLamp and its novel hybrid sensing approach. This approach combines RGB and depth input for detecting and tracking movements of multiple mobile screens with sub-centimetre precision by exploiting their optical characteristics in both the RGB and IR range.

The Hybrid Patient Record (HyPR) consists of a paper record, tablet to access the digital record and a mobile mediating sensing platform (HyPR Device) that supports pairing of the tablet and the paper record using proximity sensing, and augments the record with a notification system (color and sound) and indoor location tracking. We report on the user-centered design process and medical simulation in which the HyPR was tested by 8 clinicians.

Pre-kindergarten children have problems with double tap and long pressed gestures. We empirically test specific strategies to deal with these issues. The study shows that the implementation of these design guidelines has a positive effect on success rates, becoming feasible their inclusion in future touch based applications for pre-kindergarten children.

Even though Multi Surface Environments (MSE) and how to perform interactions in these environments have received much attention during recent years, interaction with geospatial data in MSEs is still limited. To summarize the earlier research in this area, this paper presents a systematic review of MSE interactions with geospatial data.

We developed a novel tactile display that provides a multi-push sensation by using a 6-axis motion platform. When a second finger touches the display, the surface panel declines and the position of the first finger remains unchanged. Thus, the user can push at two positions on the surface panel.

NetBoards are situated displays designed to fulfil and augment the role of non-digital personal noticeboards in the workplace. By replacing these with large, networked,touch-enabled displays, we replicate the existing physical systems' flexibility and ease-of-use, while enabling more expressive content creation techniques and remote connectivity.

We propose P.I.A.N.O., a piano learning system with interactive projection that facilitates a fast learning process. Note information in form of an enhanced piano roll notation is directly projected onto the instrument and allows mapping of notes to piano keys without prior sight-reading skills. We report the results of two user studies, which show that P.I.A.N.O. supports faster learning.

PolyChrome is an application framework for creating collaborative web visualization. The framework supports (1) synchronous and asynchronous collaboration through an API for operation distribution and display management; (2) co-browsing existing webpages; and (3) maintenance of state and input events for managing consistency and conflicts. PolyChrome makes way for creating multi-device ecosystems for ubiquitous analytics over the web.

We present SleeD, a touch-sensitive Sleeve Display that facilitates interaction with multi-touch display walls. We discuss design implications and propose techniques for user-specific interfaces, personal views and data transfer. We verified our concepts using two prototypes, several example applications and an observational study.

This work presents pen-based techniques to annotate maps and selectively convert those annotations into spatial queries to search for POIs within explicitly input or implicit scopes (areas and paths). We further provide pen gestures to calculate routes between multiple locations. A user evaluation confirms the potential of our techniques.

We present the first structure editor of handwritten mathematics oriented to the calculational mathematics involved in algorithmic problem solving. The editor provides features that allow reliable structure manipulation of mathematical content, including structured selection of expressions, the use of gestures to manipulate formulae, and definition/redefinition of operators in runtime.

In collaborative complex domains, while automation can help manage complex tasks and rapidly update information, operators may be unable to keep up with the dynamic changes. To improve situation awareness in co-located environments on digital tabletop computers, we developed an interactive event timeline that enables exploration of historical system events, using a collaborative digital board game as a case study.

Surface Ghosts virtual embodiments provide user feedback during Pick-and-Drop style cross-device transfer in a tabletop-centric multi-surface environment. Surface Ghosts take the form of semi-transparent “ghosts” of transferred objects displayed under the “owner’s” hand on the tabletop surface during transfer.

We report on two studies to understand the effect of view techniques on collaboration and awareness. We first studied the performance of view-techniques on spatial arrangement to understand by how much different view configuration affect group performance in mixed-focus tasks. Then we studies how do people choose to configure their environment when working in a realistic mixed-focus activity.

This paper describes a study that has been conducted to evaluate selected fun.tast.tisch. (a tabletop system for neuro-rehabilitation) modules. The target group usually involves patients (most of them incurred acquired brain injury) and therapists. The study described here focuses on the usability of the system from therapists’ point of view.

This paper identifies an emerging trend of technical research aimed at the "digitization of architectural space", and describes the range of contributions the HCI community can make in this domain. Opens up a new research field with relevance not only to HCI but also to architecture and urban design.

We present freeform interactive loudspeakers for creating spatial sound experiences from a variety of surfaces. Our proposed system creates directional sound and can be easily embedded into architecture, furniture and many everyday objects.

We present a video-based gesture dataset, a taxonomy, and a methodology for annotating video-based gesture datasets. We design and use a crowd-sourced classification task to annotate the videos. The results are made available through a web-based visualization that allows researchers and designers to explore the dataset.

This paper replicates Morris's Web on the Wall guessability study first using Wizard of Oz for multimodal interaction elicitation around Kinect. To obtain reproducible and implementable user-defined interaction sets, the paper argues for extending the methodology to include a draft of the system for mixed-initiative elicitation with real system dialogue.