Andreas Shepley

Andrew is a lecturer and researcher at the University of New England (UNE). He is an expert in deep learning, innovating at the algorithmic and architectural level, with extensive experience in the application of deep learning methods in the wild..

Andrew obtained his PhD in Artificial Intelligence from UNE, where he investigated contemporary challenges in adoption of computer vision technologies in applied sciences. He proposed Confluence, a novel algorithm that improves accuracy and recall of object detections, achieving state-of-the-art results in high occlusion settings. He published effective strategies to better fine-tune computer vision models, improving robustness and domain invariance in automated ecological image processing. He incepted and developed U-Infuse, a software application aimed at democratising access to deep learning model development and deployment. Andrew has a BSc. Hons from the University of New South Wales (UNSW), a Grad. Degree in Information Science (UNSW), and a Grad Dip. Education (UNE).

Andrew is the creator and UC of Deep Learning (COSC351/551) and Object Oriented Programming (COSC120). He is currently researching the application of transformers to market volatility prediction and leading the development of platform-agnostic, offline, on-device deep learning software. He is open to supervise HDR students in the field of artificial intelligence.