Pose Estimation from RGB Images of Highly Symmetric Objects using a Novel Multi-Pose Loss and Differential Rendering
Stefan Hein Bengtson,Hampus Åström,Thomas B. Moeslund,Elin A. Topp,Volker Krueger,Stefan Hein Bengtson,Hampus Åström,Thomas B. Moeslund,Elin A. Topp,Volker Krueger
We propose a novel multi-pose loss function to train a neural network for 6D pose estimation, using synthetic data and evaluating it on real images. Our loss is inspired by the VSD (Visible Surface Discrepancy) metric and relies on a differentiable renderer and CAD models. This novel multi-pose approach produces multiple weighted pose estimates to avoid getting stuck in local minima. Our method re...
