Deep Virtual Markers for Articulated 3D Shapes
This work addresses the need for accurate and dense marker estimation in 3D articulated models, which is incremental as it builds on existing methods with novel enhancements for specific bottlenecks.
The paper tackles the problem of estimating dense positional information for 3D articulated shapes like humans by proposing deep virtual markers, a framework that maps 3D points into virtual marker labels using a sparse convolutional neural network with soft labels based on geodesic distance. The result outperforms state-of-the-art on the FAUST challenge and shows strong generalizability across unseen data and different 3D data types.
We propose deep virtual markers, a framework for estimating dense and accurate positional information for various types of 3D data. We design a concept and construct a framework that maps 3D points of 3D articulated models, like humans, into virtual marker labels. To realize the framework, we adopt a sparse convolutional neural network and classify 3D points of an articulated model into virtual marker labels. We propose to use soft labels for the classifier to learn rich and dense interclass relationships based on geodesic distance. To measure the localization accuracy of the virtual markers, we test FAUST challenge, and our result outperforms the state-of-the-art. We also observe outstanding performance on the generalizability test, unseen data evaluation, and different 3D data types (meshes and depth maps). We show additional applications using the estimated virtual markers, such as non-rigid registration, texture transfer, and realtime dense marker prediction from depth maps.