MaskedFusion: Mask-based 6D Object Pose Estimation

MaskedFusion is a framework to estimate the 6D pose of objects using RGB-Ddata, with an architecture that leverages multiple sub-tasks in a pipeline toachieve accurate 6D poses. 6D pose estimation is an open challenge due tocomplex world objects and many possible problems when capturing data from thereal world, e.g., occlusions, truncations, and noise in the data. Achievingaccurate 6D poses will improve results in other open problems like robotgrasping or positioning objects in augmented reality. MaskedFusion improves thestate-of-the-art by using object masks to eliminate non-relevant data. With theinclusion of the masks on the neural network that estimates the 6D pose of anobject we also have features that represent the object shape. MaskedFusion is amodular pipeline where each sub-task can have different methods that achievethe objective. MaskedFusion achieved 97.3% on average using the ADD metric onthe LineMOD dataset and 93.3% using the ADD-S AUC metric on YCB-Video Dataset,which is an improvement, compared to the state-of-the-art methods. The code isavailable on GitHub (https://github.com/kroglice/MaskedFusion).