- Room No: G.1.31
- Telephone: +64 7 838 4466
- Extension: 6729
- Facsimile: +64 7 838 4155
BCMS Honours Topic: "Feature learning with Convolutional Neural Networks."
Feature Learning is the process of learning to transform a complex feature vector, such as an image, to a compact and efficient representation that can be used by Machine Learning algorithms. In the field of Computer Vision, Convolutional Neural Networks are the state of the art method for image classification due to their feature learning abilities.
The two main goals of the project are to accelerate the speed at which Convolutional Neural Networks are trained, and to increase the usefulness of the features they learn. The project will be evaluated based on the overall speedup of the time taken to learn features, and the accuracy achieved on several Computer Vision and Artificial Intelligence tasks.