I have been working on computer vision and 3d geometry since 2013, and deep learning since 2015. My focus has primarily been in the field of aerial imagery. At Verisk I designed and produced a machine learning pipeline all the way from labeling data, developing models, and deploying at scale an api for user requests. I have developed models for the following use cases:- Instance segmentation for large scale building detection- Object detection networks- Semantic segmentation for various features related to structures and properties- Classification networks- 3d convolutional neural networks for processing pointcloud dataIn addition to neural networks I have worked extensively with 3d geometry problems.As part of my work and research I have been awarded the following patents:https://patents.google.com/patent/AU2022275479A1https://patents.google.com/patent/US10909757B2https://patents.google.com/patent/US11062166B2https://patents.google.com/patent/US11094113B2https://patents.google.com/patent/WO2022159592A1https://patents.google.com/patent/US20220366646A1https://patents.google.com/patent/US20210398347A1https://patents.google.com/patent/AU2021296852A1https://patents.google.com/patent/US20220327722A1