T
Toby P. Breckon
Researcher at Durham University
Publications - 238
Citations - 6704
Toby P. Breckon is an academic researcher from Durham University. The author has contributed to research in topics: Object detection & Convolutional neural network. The author has an hindex of 34, co-authored 221 publications receiving 4692 citations. Previous affiliations of Toby P. Breckon include University of Edinburgh & Cranfield University.
Papers
More filters
Book ChapterDOI
GANomaly : semi-supervised anomaly detection via adversarial training.
TL;DR: In this paper, a conditional generative adversarial network (GAN) is used for anomaly detection in a one-class, semi-supervised learning paradigm, where an encoder-decoder-encoder sub-network is employed to map the input image to a lower dimension vector, which is then used to reconstruct the generated output image.
Book
Fundamentals of Digital Image Processing: A Practical Approach with Examples in Matlab
TL;DR: This is an introductory to intermediate level text on the science of image processing, which employs the Matlab programming language to illustrate some of the elementary, key concepts in modern image processing and pattern recognition.
Proceedings ArticleDOI
Real-Time Monocular Depth Estimation Using Synthetic Data with Domain Adaptation via Image Style Transfer
TL;DR: In this paper, the authors take advantage of image style transfer and adversarial training to predict pixel perfect depth from a single real-world color image based on training over a large corpus of synthetic environment data.
Journal ArticleDOI
Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery
TL;DR: The comparative performance of these techniques is illustrated and it is shown that object localization strategies cope well with cluttered X-ray security imagery, where classification techniques fail, and that fine-tuned CNN features yield superior performance to conventional hand-crafted features on object classification tasks within this context.
Proceedings ArticleDOI
Transfer learning using convolutional neural networks for object classification within X-ray baggage security imagery
TL;DR: A transfer learning paradigm is employed such that a pre-trained CNN, primarily trained for generalized image classification tasks where sufficient training data exists, can be specifically optimized as a later secondary process that targets specific this application domain.