Accurate tree segmentation is a key step in extracting individual tree metrics from forest laser scans, and is essential to understanding ecosystem functions in carbon cycling and beyond. Over the past decade, tree segmentation algorithms have advanced rapidly due to developments in AI. However, existing public 3D forest datasets are not large enough to build robust tree segmentation systems. Motivated by the success of synthetic data in other domains such as self-driving, we investigate whether similar approaches can help with tree segmentation. In place of expensive field data collection and annotation, we use synthetic data during pretraining, and then require only minimal, real forest plot annotation for fine-tuning. We have developed Cambridge Arboreal Modelling Panoptic 3D (CAMP3D), a new synthetic data generation pipeline to do this for forest vision tasks, integrating advances in game engines with physics-based LiDAR simulation. Using CAMP3D, we have produced a comprehensive, diverse, annotated 3D forest dataset on an unprecedented scale. Extensive experiments with a state-of-the-art tree segmentation algorithm and a popular real dataset show that our synthetic data can substantially reduce the need for labelled real data. After fine-tuning on just a single, real, forest plot of less than 0.1 hectare, the pretrained model achieves segmentations that are competitive with a model trained on the full scale real data. We have also identified critical factors for successful use of synthetic data: physics, diversity, and scale, paving the way for more robust 3D forest vision systems in the future. Our CAMP3D pipeline and the resulting dataset are available at https://github.com/yihshe/CAMP3D.git.

Videoconferencing has become a ubiquitous medium for collaborative work. It does suffer however from various drawbacks such as zoom fatigue. This paper addresses the quality of user experience by exploring an enhanced system concept with the capability of conveying gaze and attention. Gazing Heads is a round-table virtual meeting concept that uses only a single screen per participant. It enables direct eye contact, and signals gaze via controlled head rotation. The technology to realise this novel concept is not quite mature though, so we built a camera-based simulation for four simultaneous videoconference users. We conducted a user study comparing Gazing Heads with a conventional “Tiled View” video conferencing system, for 20 groups of 4 people, on each of two tasks. The study found that head rotation clearly conveys gaze and strongly enhances the perception of attention. Measurements of turn-taking behaviour did not differ decisively between the two systems (though there were significant differences between the two tasks). A novel insight in comparison to prior studies is that there was a significant increase in mutual eye contact with Gazing Heads, and that users clearly felt more engaged, encouraged to participate and more socially present. Overall, participants expressed a clear preference for Gazing Heads. These results suggest that fully implementing the Gazing Heads concept, using modern computer vision technology as it matures, could significantly enhance the experience of videoconferencing.