Lukas Winiwarter of the 3DGeo group was invited by the Austrian Society of Surveying and Geoinformation (OVG) to give a talk on the application of deep learning on point clouds, which took place on March 24. In his talk, Lukas presented four different state-of-the-art approaches to consider the irregular, unordered structure of point clouds, which allow a deep neural network to learn a representation of the neighbourhood of each point.
Some impressions from the talk (in German):
After presenting to almost 70 participants, a lively discussion showed the relevance of this topic in academia, public administration, and in the private sector. A big challenge in deep learning is the requirement for vast amounts training data, which may in the future be partially provided by simulation – e.g. using our own open-source general-purpose LiDAR simulator, HELIOS++.
We thank the OVG for the organization of the talk and are looking forward to future events.