Machine Learning to Model Complex Natural Phenomena through Computer Vision Computer Graphics Applications

Computer Graphics is used by artists and animators ubiquitously across many industries, including but not limited to films and games. The users are story-tellers, more interested in creating a narrative that adhering to the laws of physics. This makes a real problem for the engineers that work with them: physics is the obvious way to model real world things, but the artists require the models to breach physics as and when required by the narrative. Moreover, physical models are difficult to construct and control: artists want intuitive models, not a host of buttons and levers that mean little to them. These problems are particularly pronounced in the case of objects such as fire, water, smoke, trees, and a host of other associated complex natural phenomena.

This project builds on successful work already conducted at Bath. The idea is to acquire three dimensional dynamic models of natural phenomena using video as input. The output model will be in such a form as to not only reproduce the appearance of physical reality by default but also to provide intuitive modes for artists to edit both appearance and motion, and to create new models using existing models as examples.

You will use make use of the latest developments in machine learning to invent new computer vision algorithms that learn to see the phenomena, and by seeing produce a model that conforms far better to the needs of users. You will be expected to publish in the highest quality journals and conferences, and you will travel internationally to present your work. Where applicable, you will work with industry to deploy your inventions into real systems. On completion of this PhD you will be an expert in a position to pursue either an academic or an industrial career.

Research Environment:

You will join the Visual Computing group at Bath. The group enjoys an international reputation: we regularly publish in all of the top-ranking Computer Vision and Computer Graphics conferences and journals. We have many international collaborations, including with the Max-Plank Institute in Germany, and the elite Tsinghua University in China.

The group plays host to the CAMERA project, which provides state of the art, industrial quality facilities for motion capture; CAMERA works with many companies in the creative sector. The group also runs the Centre for Digital Entertainment, which places EngD level students into industry for three year, typically but not exclusively in the creative sector.

Candidate:

The successful candidate will have an excellent first degree in a highly numerate discipline. An enthusiasm for research and exploration is essential.

Applications:

Informal enquiries should be directed to Professor Peter Hall, [Email Address Removed].

Formal applications should be made via the University of Bath’s online application form:
https://samis.bath.ac.uk/urd/sits.urd/run/siw_ipp_lgn.login?process=siw_ipp_app&code1=RDUCM-FP01&code2=0013

Please ensure that your quote the supervisor’s name and project title in the ‘Your research interests’ section.

More information about applying for a PhD at Bath may be found here:
http://www.bath.ac.uk/guides/how-to-apply-for-doctoral-study/

The start date will be no later than the end of March 2019.