Our workshop on Scene Understanding and Semantic SLAM: Progress, Applications, and Limitations will discuss the current progress, applications, and limitations of robotic scene understanding and semantic simultaneous localization and mapping (SLAM). We are motivated by the quickly accelerating appearance new research outputs that investigate how classical SLAM techniques and deep-learning based visual object detection or segmentation can be combined in innovative ways, and used to support scene understanding, navigation, and manipulation.


The Importance of Uncertainty in Deep Learning for Robotics was the topic of our workshop at IROS 2019. Invited speakers and contributed papers discussed the importance of uncertainty in deep learning for robotic perception, but also action. We publicly presented our work towards a new Scene Understanding challenge for the first time.

We are organised a competition and workshop at CVPR 2019, introducing the probabilistic object detection challenge. Four participating teams presented their approaches and results. More details and links to their papers can be found on the workshop website.