Approximate Computation to Minimise Footprint

Back to search results
The research in this area is to develop simpler but reliable computing systems that can run on low-cost, small-scale platforms, for use in applications such as image recognition and machine learning. Cheaper reliable systems would lead to wider availability of this type of technology to end users. This will be achieved by developing innovative approximate algorithmic and circuit techniques. This project will enable low cost systems which are reliable, dependable and sufficiently accurate. This project will result in publications, and new intellectual property (such as patents and inventions).

The ideal candidate will be a student who is well versed in the areas of digital circuit design, computer architecture and software design. The candidate should ideally have an interest in systems design and be interested in low power methodologies.

Supervisory team
Sri
Parameswaran

Engineering
Computer Science and Engineering
Aleksandar
Ignjatovic

Engineering
Computer Science and Engineering
Angelo
Ambrose

Seeing Machines Pty Ltd
Other
sri.parameswaran@unsw.edu.au