My current research revolves around machine vision for applications in agriculture and robotics. My primary focus is making networks more general for cross-domain accuracy and segmentation of objects in unseen domains through multi-task learning. This includes weeding in arable farmland with species-level classification and crop segmentation and their quality in glasshouses. I also have a focus on making algorithms agnostic to both the field they are operating and the platform they have been deployed on. These approaches will aim to create better monitoring of crop/fruit while enabling key decision-making processes for the end-users.