# EPSilon: Efficient Point Sampling for Lightening of Hybrid-based 3D Avatar Generation

arXiv: https://arxiv.org/abs/2507.13648 · HTML: https://mlsanity.com/paper/2507.13648/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 35 of its 41 references that are in the map.
- First posted on arXiv 18 Jul 2025.

## Builds on (most cited first, 30 of 35)

- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Instant Neural Graphics Primitives with a Multiresolution Hash Encoding](https://mlsanity.com/paper/2201.05989/) (2022)
- [Expressive Body Capture: 3D Hands, Face, and Body from a Single Image](https://mlsanity.com/paper/1904.05866/) (2019)
- [D-NeRF: Neural Radiance Fields for Dynamic Scenes](https://mlsanity.com/paper/2011.13961/) (2020)
- [TensoRF: Tensorial Radiance Fields](https://mlsanity.com/paper/2203.09517/) (2022)
- [Nerfies: Deformable Neural Radiance Fields](https://mlsanity.com/paper/2011.12948/) (2020)
- [Point-NeRF: Point-based Neural Radiance Fields](https://mlsanity.com/paper/2201.08845/) (2022)
- [Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies](https://mlsanity.com/paper/2105.02872/) (2021)
- [ARCH: Animatable Reconstruction of Clothed Humans](https://mlsanity.com/paper/2004.04572/) (2020)
- [PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop](https://mlsanity.com/paper/2103.16507/) (2021)
- [Video Based Reconstruction of 3D People Models](https://mlsanity.com/paper/1803.04758/) (2018)
- [SNARF: Differentiable Forward Skinning for Animating Non-Rigid Neural Implicit Shapes](https://mlsanity.com/paper/2104.03953/) (2021)
- [ARCH++: Animation-Ready Clothed Human Reconstruction Revisited](https://mlsanity.com/paper/2108.07845/) (2021)
- [SelfRecon: Self Reconstruction Your Digital Avatar from Monocular Video](https://mlsanity.com/paper/2201.12792/) (2022)
- [ARAH: Animatable Volume Rendering of Articulated Human SDFs](https://mlsanity.com/paper/2210.10036/) (2022)
- [InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds](https://mlsanity.com/paper/2212.10550/) (2022)
- [Collaborative Regression of Expressive Bodies using Moderation](https://mlsanity.com/paper/2105.05301/) (2021)
- [VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids](https://mlsanity.com/paper/2206.07695/) (2022)
- [Detailed Human Avatars from Monocular Video](https://mlsanity.com/paper/1808.01338/) (2018)
- [Animatable Neural Radiance Fields from Monocular RGB Videos](https://mlsanity.com/paper/2106.13629/) (2021)
- [NerfAcc: Efficient Sampling Accelerates NeRFs](https://mlsanity.com/paper/2305.04966/) (2023)
- [Capturing and Animation of Body and Clothing from Monocular Video](https://mlsanity.com/paper/2210.01868/) (2022)
- [NerfAcc: A General NeRF Acceleration Toolbox](https://mlsanity.com/paper/2210.04847/) (2022)
- [AdaNeRF: Adaptive Sampling for Real-time Rendering of Neural Radiance Fields](https://mlsanity.com/paper/2207.10312/) (2022)
- [X-Avatar: Expressive Human Avatars](https://mlsanity.com/paper/2303.04805/) (2023)
- [Fast-SNARF: A Fast Deformer for Articulated Neural Fields](https://mlsanity.com/paper/2211.15601/) (2022)
- [AniFaceGAN: Animatable 3D-Aware Face Image Generation for Video Avatars](https://mlsanity.com/paper/2210.06465/) (2022)
- [Learning Disentangled Avatars with Hybrid 3D Representations](https://mlsanity.com/paper/2309.06441/) (2023)
- [Fast Learning Radiance Fields by Shooting Much Fewer Rays](https://mlsanity.com/paper/2208.06821/) (2022)
- [OmniAvatar: Geometry-Guided Controllable 3D Head Synthesis](https://mlsanity.com/paper/2303.15539/) (2023)

## Abstract (arXiv)

The rapid advancement of neural radiance fields (NeRF) has paved the way to generate animatable human avatars from a monocular video. However, the sole usage of NeRF suffers from a lack of details, which results in the emergence of hybrid representation that utilizes SMPL-based mesh together with NeRF representation. While hybrid-based models show photo-realistic human avatar generation qualities, they suffer from extremely slow inference due to their deformation scheme: to be aligned with the mesh, hybrid-based models use the deformation based on SMPL skinning weights, which needs high computational costs on each sampled point. We observe that since most of the sampled points are located in empty space, they do not affect the generation quality but result in inference latency with deformation. In light of this observation, we propose EPSilon, a hybrid-based 3D avatar generation scheme with novel efficient point sampling strategies that boost both training and inference. In EPSilon, we propose two methods to omit empty points at rendering; empty ray omission (ERO) and empty interval omission (EIO). In ERO, we wipe out rays that progress through the empty space. Then, EIO narrows down the sampling interval on the ray, which wipes out the region not occupied by either clothes or mesh. The delicate sampling scheme of EPSilon enables not only great computational cost reduction during deformation but also the designation of the important regions to be sampled, which enables a single-stage NeRF structure without hierarchical sampling. Compared to existing methods, EPSilon maintains the generation quality while using only 3.9% of sampled points and achieves around 20 times faster inference, together with 4 times faster training convergence. We provide video results on https://github.com/seungjun-moon/epsilon.