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

We propose a novel neural rendering pipeline, Hybrid Volumetric-TexturalRendering (HVTR), which synthesizes virtual human avatars from arbitrary posesefficiently and at high quality. First, we learn to encode articulated humanmotions on a dense UV manifold of the human body surface. To handle complicatedmotions (e.g., self-occlusions), we then leverage the encoded information onthe UV manifold to construct a 3D volumetric representation based on a dynamicpose-conditioned neural radiance field. While this allows us to represent 3Dgeometry with changing topology, volumetric rendering is computationally heavy.Hence we employ only a rough volumetric representation using a pose-conditioneddownsampled neural radiance field (PD-NeRF), which we can render efficiently atlow resolutions. In addition, we learn 2D textural features that are fused withrendered volumetric features in image space. The key advantage of our approachis that we can then convert the fused features into a high-resolution,high-quality avatar by a fast GAN-based textural renderer. We demonstrate thathybrid rendering enables HVTR to handle complicated motions, renderhigh-quality avatars under user-controlled poses/shapes and even looseclothing, and most importantly, be efficient at inference time. Ourexperimental results also demonstrate state-of-the-art quantitative results.

HVTR: Hybrid Volumetric-Textural Rendering for Human Avatars