MUT - 1 New generation context-aware human computer interface
NICTA Project - Dr Ronnie Taib (ronnie.taib@nicta.com.au)
The context in which applications (e.g. web pages) are used is subject to continuous micro changes (from current time to stress level) that are currently not utilised to optimise the user's experience. This exciting software project will produce an application prototype (mostly coded in Java) where such variations can be detected and processed in order to customise information output accordingly.
The project can be developed along two directions, according to the student's skills, but is also good for a team:
- Human-Computer Interaction: customisation of both the content and layout of the output;
- (Optional) Machine learning: detect, classify and interpret incoming data.
The students will flesh out a real prototype for an extremely hot issue where global research is still at the level of ideas. The prototype may also feature some state-of-the-art multimodal interaction (e.g. speech and gesture interaction), as used by the Nicta team. Potential outcomes of the research include a Java prototype, possibly with some server-side (PHP?, Java?) demonstration application, user evaluations of the prototype and associated documentation, review of competing technologies.
Students will work in a state-of-the-art research environment under guidance by experienced NICTA engineers and researchers. They will deal with camera-based gesture recognition, speech processing and multimodal experts. They will acquire professional research skills. Requirements for the project include excellent Java skills, HCI course and AI/ML course is a plus.
For further information, please contact the supervisor or consider
MUT - 2 GestureCam: An FPGA-based Smart Camera for Multimodal Human Computer Interface
NICTA Project - Dr David Shi (david.shi@nicta.com.au)
GestureCam is a smart video camera that can detect, track and recognise simple human head and hand gestures. It is built using a CMOS image sensor and a DSP-friendly FPGA. GestureCam can have many exciting applications such as human computer interface, surveillance, games and etc. FPGA based smart camera is an example of leading-edge embedded multimedia system. This project aims at implementing image/video processing algorithms and pattern recognition techniques in VHDL, and complete the prototyping of GestureCam for a multimodal web browser application. Main tasks include software development in C and VHDL, FPGA implementation and debugging, coding of a web browser extension.
Using gestures to control web and information navigation is a new idea which may lead to brand new human computer interaction technologies and applications. This project will make significant contribution to the implementation and possible commercialisation of the idea. Skill Prerequisites include a basic knowledge of image processing and experience in VHDL, FPGA, C/C++ and Java. A knowledge of computer vision and human computer interfaces is advantageous.
Expected outcomes of the research include a smart camera prototype, a demonstration of a gesture-controlled web browser using the camera and technical report and seminar. Working alongside NICTA senior researchers and engineers, the prospective student will have a great opportunity to gain skills in advanced multimedia, computer vision and human computer interaction technologies.
For further information, please contact the supervisor.
MUT - 3 New Tools for Awareness in Multi-site Computer Supported Cooperative Work
NICTA Project - Dr Julien Epps, julien.epps@nicta.com.au
Due to the growing availability and reliability of high speed networks and the increasing emphasis on computer-based collaboration tools, we will soon witness an explosion in tools for real time remote collaboration, where teams at different sites can work together on the same (or multiple) shared application in an intense manner.
This project looks at futuristic awareness tools for remote synchronous collaboration, for example extracting the outlines of user's hands over a shared tabletop display, sending these to remote sites and rendering them over the display at the remote site. The project will involve hands-on development plus an opportunity to experiment with and perform research into how these tools can improve awareness between sites and productivity of distributed teams.
Outcomes from the project will contribute to future commercialization efforts, academic publications and/or open source initiatives, and will form part of the new Braccetto project between NICTA, CSIRO and DSTO. Research results will help to answer some of the many open questions about how people will interact with each other and with computing devices in the future.
Working alongside experienced researchers and engineers, students will be able to apply their knowledge in computing and engineering, and learn more about human-computer interaction, programming, multimodal interfaces, signal processing, and research methods.
For further information, please contact the supervisor, or consider http://nicta.com.au/director/research/programs/imagen/people/julien_epps.cfm
MUT - 4 Collecting 'Cognitive Dust'
NICTA Project - Dr Julien Epps, julien.epps@nicta.com.au
An area receiving a great deal of recent interest in the areas of both co-located and distributed collaboration has been the collation of multimodal meeting data and, to a limited extent, the real time feedback of information about meetings to their participants.
This project looks at futuristic tools for monitoring different aspects of meetings using a large number of audio, video and other sensors. The term 'cognitive dust' has been coined to describe the pieces of information about meeting participants that can be captured in this manner. There are several challenges in this work: capturing the information (i.e. data acquisition and processing), representing the information (i.e. converting it into statistical or semantic information and storing it in a meaningful structure), and reasoning about the information (i.e. trying to infer what is actually going on from the partial information).
A possible outcome of the project could be a panel of visual cues that tell meeting participants how long they have each been speaking, and how long and where they have been looking during the meeting. An alternative outcome could be a tool that can summarise a meeting into its major segments and automatically produce searchable metadata that describes each of these.
Outcomes from the project will contribute to future commercialization efforts, academic publications and/or open source initiatives, and will form part of the new Braccetto project between NICTA, CSIRO and DSTO. Research results will help to answer some of the many open questions about how people will interact with each other and with computing devices in the future.
Working alongside experienced researchers and engineers, students will be able to apply their knowledge in computing and engineering, and learn more about human-computer interaction, programming, multimodal interfaces, signal processing, and research methods.
For further information, please contact the supervisor, or consider http://nicta.com.au/director/research/programs/imagen/people/julien_epps.cfm
MUT - 5 Applications for Future Groupware
NICTA Project - Dr Julien Epps (julien.epps@nicta.com.au)
Due to the growing availability and reliability of high speed networks and the increasing emphasis on computer-based collaboration tools, we will soon witness an explosion in groupware for real time remote collaboration, where there is a real need for multiple users to have simultaneous control and manipulation of a shared application.
This project takes existing platforms for the development of groupware applications and develops new applications that look ahead to a world where strictly single-user interfaces may be the exception rather than the rule. Applications could be business-oriented (e.g. for live discussions of designs, documents or visual content), or leisure-oriented (e.g. multiplayer games). This topic has a research component: students will be expected to follow up their application design with a user study examining at least one detailed aspect of either how computer-supported cooperative work can be effectively supported. An alternative topic would be to join the research and development effort in creating new groupware infrastructure to support distributed teams.
Working alongside experienced researchers and engineers, students will be able to apply their knowledge in computing and engineering, and learn more about human-computer interaction, Java programming, multimodal interfaces, signal processing, research methods and commercialisation.
For further information, please contact the supervisor or consider http://nicta.com.au/director/research/programs/imagen/people/julien_epps.cfm
[Top of Page]
