This paper considers online object-level mapping using partial point-cloud observations obtained online in an unknown environment. We develop and approach for fully Convolutional Object Retrieval and Symmetry-AIded Registration (CORSAIR). Our model extends the Fully Convolutional Geometric Features (FCGF) model to learn a global object-shape embedding in addition to local point-wise features from the point-cloud observations. The global feature is used to retrieve a similar object from a category database, and the local features are used for robust pose registration between the observed and the retrieved object. Our formulation also leverages symmetries, present in the object shapes, to obtain promising local-feature pairs from different symmetry classes for matching. We present results from synthetic and real-world datasets with different object categories to verify the robustness of our method.
Given a partial point cloud observation of an object and a database of point-cloud object models from the same category as the queried object, we want to retrieve a point cloud which is most similar to the observed object and estimate its pose with respect to the object.
Approach
We design a sparse fully convolutional network to jointly regress global and local point-cloud features, which are hierarchically correlated. We extend the point-wise FCGF feature extractor with a global embedding network. We learn local point-wise features to enable robust matching and registration of point-cloud with potentially different shapes. We learn global object-level features to enable retrieval of a point cloud from the database that is similar to the query point cloud.
During inference, we align a partially observed point-cloud with a retrieved one, and exploit object symmetry to generate matching feature pairs for registration. We construct symmetry classes within an object instance based on the local features and aid the generation of promising feature pairs for robust registration.
Results
Video (1 min)
Acknowledgements
We gratefully acknowledge support from ARL DCIST CRA W911NF17-2-0181 and ONR N00014-18-1-2828.
This webpage template was borrowed from Thai Duong.
