English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

3D hand shape and pose estimation from a single depth map is a new andchallenging computer vision problem with many applications. Existing methodsaddressing it directly regress hand meshes via 2D convolutional neuralnetworks, which leads to artefacts due to perspective distortions in theimages. To address the limitations of the existing methods, we developHandVoxNet++, i.e., a voxel-based deep network with 3D and graph convolutionstrained in a fully supervised manner. The input to our network is a 3Dvoxelized-depth-map-based on the truncated signed distance function (TSDF).HandVoxNet++ relies on two hand shape representations. The first one is the 3Dvoxelized grid of hand shape, which does not preserve the mesh topology andwhich is the most accurate representation. The second representation is thehand surface that preserves the mesh topology. We combine the advantages ofboth representations by aligning the hand surface to the voxelized hand shapeeither with a new neural Graph-Convolutions-based Mesh Registration(GCN-MeshReg) or classical segment-wise Non-Rigid Gravitational Approach(NRGA++) which does not rely on training data. In extensive evaluations onthree public benchmarks, i.e., SynHand5M, depth-based HANDS19 challenge andHO-3D, the proposed HandVoxNet++ achieves state-of-the-art performance. In thisjournal extension of our previous approach presented at CVPR 2020, we gain41.09% and 13.7% higher shape alignment accuracy on SynHand5M and HANDS19datasets, respectively. Our method is ranked first on the HANDS19 challengedataset (Task 1: Depth-Based 3D Hand Pose Estimation) at the moment of thesubmission of our results to the portal in August 2020.

HandVoxNet++: 3D Hand Shape and Pose Estimation using Voxel-Based Neural Networks