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

The task of reconstructing detailed 3D human body models from images isinteresting but challenging in computer vision due to the high freedom of humanbodies. In order to tackle the problem, we propose a coarse-to-fine method toreconstruct a detailed 3D human body from multi-view images combining voxelsuper-resolution based on learning the implicit representation. Firstly, thecoarse 3D models are estimated by learning an implicit representation based onmulti-scale features which are extracted by multi-stage hourglass networks fromthe multi-view images. Then, taking the low resolution voxel grids which aregenerated by the coarse 3D models as input, the voxel super-resolution based onan implicit representation is learned through a multi-stage 3D convolutionalneural network. Finally, the refined detailed 3D human body models can beproduced by the voxel super-resolution which can preserve the details andreduce the false reconstruction of the coarse 3D models. Benefiting from theimplicit representation, the training process in our method is memory efficientand the detailed 3D human body produced by our method from multi-view images isthe continuous decision boundary with high-resolution geometry. In addition,the coarse-to-fine method based on voxel super-resolution can remove falsereconstructions and preserve the appearance details in the finalreconstruction, simultaneously. In the experiments, our method quantitativelyand qualitatively achieves the competitive 3D human body reconstructions fromimages with various poses and shapes on both the real and synthetic datasets.

Detailed 3D Human Body Reconstruction from Multi-view Images Combining Voxel Super-Resolution and Learned Implicit Representation