Academic awarded prestigious Leverhulme Trust Senior Research Fellowship.

Posted on 13 August 2019

Senior Lecturer William Smith is one of seven 'outstanding engineering researchers' to receive the award.

The Royal Academy of Engineering has awarded William Smith the Leverhulme Trust Senior Research Fellowship to further his research on creating machines with visual capabilities. Broadly, there are two ways in which we can create such machines:

1. We can "teach" them what we know about the visual world. This knowledge is encapsulated in models that have been developed by scientists and engineers over centuries to explain processes such as the reflection of light from a surface or the projection of the 3D world to a 2D image via a camera.

2. They can "learn" from data. Here, a black box machine learning algorithm (most successfully a convolutional neural network or "CNN") is trained to map inputs to desired outputs. The black box knows nothing about the nature of the data or the problem at hand and learns to solve the entire problem from scratch. To achieve this remarkable feat, CNNs require a very large number of trainable parameters and hence, to avoid over-fitting, a very large training set. Given thousands or millions of images along with the desired output label, CNNs are capable of learning state of the art performance on a wide range of vision tasks.

Over the past 5 years, the learning-based approach has come to completely dominate computer vision, providing truly remarkable performance breakthroughs. However, this performance step change has come at a cost. We have gone from having a very good understanding of approaches that don't work very well to having methods that work extremely well for reasons we don't understand.

Dr Smith's fellowship seeks to unify these two divergent approaches, to "model what we know and learn the rest". Specifically, he will develop new architectures that allow CNNs to learn from explicit models. Besides performance improvements this approach will also bring interpretability to machine learning systems since the parameters of an explicit model have physical meaning.

The Fellowships, which are supported by the Leverhulme Trust, are designed to allow scientists to focus full-time on research by covering the costs of a replacement academic to take over their teaching and administrative duties for a year.

William said:

"The Senior Research Fellowship provides relief from teaching and administration. This gives me a unique opportunity to focus solely on an exciting line of research, travel to and work with collaborators and generate follow-on funding so that my research becomes self sustaining. I am incredibly grateful to the Royal Academy of Engineering and The Leverhulme Trust for this opportunity and can't wait to get started!"