At the 5th International ICCV Workshop on Recovering 6D Pose MVTec received an award for surface-based matching this year. Our surface-based matching method performed the best in the BOP Challenge of all participating methods on multiple evaluation datasets: LM, IC-MI, TYO-L and the industry-relevant MVTec dataset ITODD.
The BOP challenge is an annual scientific contest for rigid 3D object detection in RGB-D images. It is a public benchmark that allows comparing different methods in a fair, standardized and challenging setup. This year, the challenge was on the task of 6D localization of a varying number of instances of a different number of objects in a single RGB-D image (the ViVo task).