MobileHCI 2017: Workshop on Object Recognition for Input and Mobile Interaction
This workshop is on Sept 4th in Vienna, Austria in conjunction with MobileHCI 2017.
We thank our sponsor (GaussToys).
Aula der Wissenschaft.Wollzeile 27a
A-1010 WienScience Cafe – http://www.aula-wien.at/en/aula/science_cafe
List of Accepted Workshop Papers
- Facilitating Object Detection and Recognition through Eye Gaze
- Non-Invasive Glucose Monitoring utilizing Electromagnetic Waves
- AquaCat: Radar and Machine Learning for Fluid and Powder Identification
- EchoTube: Modular, Flexible and Multi-Point Pressure Sensors using Waveguided Ultrasound
- EyeLogging: Hybrid Eye Tracker Combining Deep Learning and Crowdsourcing Approaches
- Pinpoint: Multi-Scale Gestural Interaction for AR Facilitating
- Machine learning for recognizing and controlling smart devices
Multi-host and Multi-user Peripherals: The Start of a Journey
Best paper will be invited to publish at the International Journal of Mobile Human Computer Interaction (IJMHCI).
Final Program (last update 7.30am Sept 4th)
9.00 – 9.30 Setup time (those who are speaking should please test their machines etc.)
9.30 – 9.35 Welcome (Introduction and briefly expected outcomes)
9.35 – 9.50 Nick Gillian, Google ATP, USA
9.50 – 10.05 Mike Chen, National Taiwan University
10.05-10.20 Gierad Laput, CMU, USA
10.20 – ~10.30 Q&A please visit sli.do (#sachi) to post your questions
10.30 – 11 Coffee Break
** during the coffee break some may wish to continue discussions with panel speakers
** anyone who has tested their setup should test now
11.00 – 11.10 Facilitating Object Detection and Recognition through Eye Gaze
11.10 – 11.20 AquaCat: Radar and Machine Learning for Fluid and Powder Identification
11.20 – 11.30 Pinpoint: Multi-Scale Gestural Interaction for AR Facilitating
11.30 – 11.40 Non-Invasive Glucose Monitoring utilizing Electromagnetic Waves
11.40 – 11.50 EyeLogging: Hybrid Eye Tracker Combining Deep Learning and Crowdsourcing Approaches
11.50 – 12.00 Machine learning for recognizing and controlling smart devices
12.00 – 12.10 EchoTube: Modular, Flexible and Multi-Point Pressure Sensors using Waveguided Ultrasound
12.10 – 12.30 Q&A and Discussion
12.30 – 1.30 Lunch (and demo setup and posters)
1.30 – 2.00 Demos + Posters RadarCat/GaussSense/XAI
2.00 – 2.20 Brainstorm
2.20 – 2.57 Work: Three Parallel Breakout / Work-Group Sessions
2.57 – 3.00 1 min fast foot reporting per group
3.00 – 3.30 Coffee and breakout session continues
3.30 – 4.30 Work: Continue breakout sessions
4.30 – 4.45 Reports (Goal: to produce materials we can add to a poster we will print tonight and display at the conference. inc. sketches, demos of new concepts, gap analysis, critiques, taxonomy etc. – results of the group work and more)
4.45 – 5.00 Wrap up and next steps
Optional: Group dinner (venue to find and numbers to confirm by lunchtime)
Breaks are at 10:30, 12:30 and 15:00. Full programme is here: https://mobilehci.acm.org/2017/program.html
Today, we are seeing an emergence of devices that incorporate sensing capabilities that go beyond the traditional suite of hardware (e.g., touch sensing or proximity). These devices offer more fine-grained level of contextual information, such as object recognition, and they often vary in their size, portability, embeddability, and form factor. Despite this diversity, the manifestations of these new-generation sensing approaches will inevitably unlock many of the ubiquitous, tangible, mobile, and wearable computing ecosystems that promise to improve people’s lives.
These system are brought together by a variety of technologies, including computer vision, radar (e.g., Google ATAP’s Project Soli), acoustic sensing, fiducial tagging, and in general, IoT devices embedded with computational capabilities. Such systems open up a wide-range of applications spaces and novel forms of interaction. For instance, object-based interactions offer rich, contextual information that can power a wide range of user-centric applications (e.g., factory line optimization and safety, automatic grocery checkout, new forms of tangible interactions). Where, and how these interactions are applied also adds a new dimension to these applications (e.g., if a mobile device can detect which part of your body it is tapped into, it can launch the food app when tapped to your stomach).
Although the last few years have seen an increasing amount of research in this area, knowledge about this subject remains under explored, fragmented, and cuts across a set of related but heterogeneous issues. This workshop brings together researchers and practitioners interested in the challenges posed by “Object Recognition for Input and Mobile Interaction”.
This workshop aims to bring together active and interested researchers in sensing techniques, particularly those that advance input and interaction using novel capabilities (e.g., object recognition, radar sensing) across a range of modalities (e.g., mobile devices, wearables). This workshop will foster a scholarly environment for sharing approaches and experiences, identifying research and deployment challenges, and envision the next generation of applications that rely on widely deployed sensor systems, paying close attention to not just one, but an ecosystem of touchless, interaction devices.
There are many challenges with building the underlying system infrastructures for object recognition for mobile interaction. For example, what are the standards and operating system requirements? How can vision, radar or acoustic sensing or tagging systems which exist in multiple mobile and wearable elements act in a coordinated manner to reliably determine object interaction? What types of software libraries, simulators and IDEs are required for development? How can we ensure inter-device interoperability in the face of heterogeneous device configurations and varying approaches to object recognition?
In addition, there are significant challenges in the design and deployment of these kinds of object based interfaces. For example, what are the implications of interfaces that rely on sensing the movement and type of passive, physical objects? How might multiple points of sensing on, and around the body open up new design considerations? Which elements of the interface are best distributed in outputs? What new interaction techniques are necessary in these environments? What are the performance, comfort, and preference consequences of relying of object-interaction as a mobile interface?
We invite short (upto 2 pages) and longer (upto 6 pages) contributions from researchers and practitioners working in the area of object recognition and interaction. Some examples of topics of interest can be found below. Our goal is to have a group who can share approaches and experiences, identify research and deployment challenges, and envision the next generation of applications that rely on object recognition and interaction. Such systems can be embedded in personal, wearable and mobile devices i.e. physically decoupled in different ways yet are virtually coupled due to the interactions they support. We expect researchers who attend the workshop to be working on different views of the problem (e.g., at the interaction technique, application, middleware or hardware level), with a wide range of sensing systems and interaction technologies (e.g., Project Soli, computer vision, alternative hardware, …), and in a wide range of object interaction applications (e.g., education, on-the-go interaction, medicine, smart sensing).
Send submissions to firstname.lastname@example.org before the end of May 19th (friday, 24:00 AOE, using ACM SIGCHI format. Submissions will be peer reviewed by an international program committee. Submissions do not have to be anonymous.
Topics of interest
Broadly speaking this workshop is interested in object recognition work based on computer vision, radar (e.g. Project Soli by the Google ATAP (Advanced Technology and Projects)), acoustic sensing, tagging, smart objects etc. Participants are encouraged to bring their mobile and desktop based systems suitable for object recognition for mobile interaction. The following are example topics of interest but we are not limited to these:
- Understanding the design space and identifying factors that influence user interactions in this space
- Developing evaluation strategies to cope with the complex nature of computer vision or radar-based object interaction
- Ethnography and user studies of tangible and object interaction
- Examples of applications of object interaction
- Social factors that influence the design of suitable interaction techniques for object interaction
- Exploring interaction techniques that facilitate multi-person object interaction
- Novel input mechanisms for single and multi-point sensing systems (eg. in mobile, tablet and wrist worn devices)
- SDK/APIs, IDEs, and hardware platforms for the development of object recognition.
International Program Committee
- Andrea Bianchi, KAIST, Korea
- Nicholas Gillian, ATAP Google, USA
- Chris Harrison, Carnegie Mellon University, USA
- Niels Henze, University of Stuttgart, Germany
- Gierad Laput, Carnegie Mellon University, USA
- Rong-Hao Liang, Eindhoven University of Technology, Netherlands
- Aaron Quigley, University of St Andrews, UK
- Kaisa Vaananen, Tampere University of Technology, Finland
- Hui-Shyong Yeo, University of St Andrews, UK
- Liwei Chan, National Chiao Tung University, Taiwan
- Cheng Zhang, Georgia Tech, USA
- Chouchang (Jack) Yang, Disney Research
Last updated August 24th.