RECORD
DRaCoN -- Differentiable Rasterization Conditioned Neural Radiance Fields for Articulated Avatars
Computer Vision and Pattern Recognition · Submitted 29 Mar 2022
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Experiments on the challenging ZJU-MoCap and Human3.6M datasets indicate that DRaCoN outperforms state-of-the-art methods both in terms of error metrics and visual quality.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Acquisition and creation of digital human avatars is an important problem with applications to virtual telepresence, gaming, and human modeling. Most contemporary approaches for avatar generation can be viewed either as 3D-based methods, which use multi-view data to learn a 3D representation with appearance (such as a mesh, implicit surface, or volume), or 2D-based methods which learn photo-realistic renderings of avatars but lack accurate 3D representations. In this work, we present, DRaCoN, a framework for learning full-body volumetric avatars which exploits the advantages of both the 2D and 3D neural rendering techniques. It consists of a Differentiable Rasterization module, DiffRas, that synthesizes a low-resolution version of the target image along with additional latent features guided by a parametric body model. The output of DiffRas is then used as conditioning to our conditional neural 3D representation module (c-NeRF) which generates the final high-res image along with body geometry using volumetric rendering. While DiffRas helps in obtaining photo-realistic image quality, c-NeRF, which employs signed distance fields (SDF) for 3D representations, helps to obtain fine 3D geometric details. Experiments on the challenging ZJU-MoCap and Human3.6M datasets indicate that DRaCoN outperforms state-of-the-art methods both in terms of error metrics and visual quality.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Acquisition and creation of digital human avatars is an important problem with applications to virtual telepresence, gaming, and human modeling. | background 1.00 | 0.03 | 0.02 | asserted 1.00 |
| s1 | Most contemporary approaches for avatar generation can be viewed either as 3D-based methods, which use multi-view data to learn a 3D representation with appearance (such as a mesh, implicit surface, or volume), or 2D-based methods which learn photo-realistic renderings of avatars but lack accurate 3D representations. | background 0.95 | 0.05 | 0.02 | asserted 0.98 |
| s2 | In this work, we present, DRaCoN, a framework for learning full-body volumetric avatars which exploits the advantages of both the 2D and 3D neural rendering techniques. | contribution 1.00 | 0.89 | 0.12 | asserted 0.99 |
| s3 | It consists of a Differentiable Rasterization module, DiffRas, that synthesizes a low-resolution version of the target image along with additional latent features guided by a parametric body model. | method 0.98 | 0.86 | 0.03 | asserted 1.00 |
| s4 | The output of DiffRas is then used as conditioning to our conditional neural 3D representation module (c-NeRF) which generates the final high-res image along with body geometry using volumetric rendering. | method 1.00 | 0.84 | 0.04 | asserted 1.00 |
| s5 | While DiffRas helps in obtaining photo-realistic image quality, c-NeRF, which employs signed distance fields (SDF) for 3D representations, helps to obtain fine 3D geometric details. | method 0.64 | 0.62 | 0.08 | asserted 0.99 |
| s6 | Experiments on the challenging ZJU-MoCap and Human3.6M datasets indicate that DRaCoN outperforms state-of-the-art methods both in terms of error metrics and visual quality. | result 1.00 | 0.96 | 0.85 | hedged possibility 0.62 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Acquisition and creation of digital human avatars is an important problem with applications to virtual telepresence, gaming, and human modeling. Most contemporary approaches for avatar generation can be viewed either as 3D-based methods, which use multi-view data to learn a 3D representation with appearance (such as a mesh, implicit surface, or volume), or 2D-based methods which learn photo-realistic renderings of avatars but lack accurate 3D representations. In this work, we present, DRaCoN, a framework for learning full-body volumetric avatars which exploits the advantages of both the 2D and 3D neural rendering techniques. It consists of a Differentiable Rasterization module, DiffRas, that synthesizes a low-resolution version of the target image along with additional latent features guided by a parametric body model. The output of DiffRas is then used as conditioning to our conditional neural 3D representation module (c-NeRF) which generates the final high-res image along with body geometry using volumetric rendering. While DiffRas helps in obtaining photo-realistic image quality, c-NeRF, which employs signed distance fields (SDF) for 3D representations, helps to obtain fine 3D geometric details. Experiments on the challenging ZJU-MoCap and Human3.6M datasets indicate that DRaCoN outperforms state-of-the-art methods both in terms of error metrics and visual quality.
Human review
How the literature uses it
3 arXiv papers cite this in 4 sentences
Papers citing it · 2 sentences without a confident label
Cites it as background 2 papers
-
Several recent methods study the problem of composing scene elements with implicit representation [ref] , which have gained a great success in modeling scenes.
LANe: Lighting-Aware Neural Fields for Compositional Scene Synthesis · Implicit 3D methods · 0.93
-
Reconstructing 3D humans from videos is a fast growing field, and there are many concurrent works [ref] .
Animatable Implicit Neural Representations for Creating Realistic Avatars from Videos · Related work · 1.00
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.