AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Animatable Gaussian human avatars

Builds deformable Gaussian representations of humans and hands for real-time avatar rendering.

42 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2024–2026) · an approach in Gaussian splatting, keyword cluster “3d gaussian · gaussian splatting · splatting 3dgs”

State History Tried Checklist Preprints Who Export

Where it stands

Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.

11 papers in 2023–24 and 31 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 9.0% to 4.3%.

How this is decided

A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.

How it got here

2016: 0162017: 0172018: 0182019: 0192020: 0202021: 0212022: 0222023: 0232024: 11242025: 19252026: 1226

Most cited paper of each year

Earliest papers here

What has been tried

30 papers here report 275 measured changes (a component added, removed or swapped, compared with a baseline): helped 85%, no clear effect 2%, hurt 12%, mixed 1%. Papers mostly publish changes that helped, so "helped" is over-represented.

Ablations in recent papers

What was changedMeasured onResultEffectPaper
Remove color MLPZJU-MoCap · LPIPS29.58 -> 31.24Hurt3DGS-Avatar: Animatable Avatars via Deformable 3D Gaussian Splatting (CVPR 2024 · 166 citations)
add pose refinementZJU_MoCap (386 sequence, masked area) · PSNR27.94 -> 28.08HelpedGauHuman: Articulated Gaussian Splatting from Monocular Human Videos (CVPR 2024 · 114 citations)
add learned per-Gaussian skinning weights instead of using closest-vertex SMPL weightsTHuman4 novel views · PSNR30.97 -> 31.99HelpedHuman Gaussian Splatting: Real-time Rendering of Animatable Avatars (CVPR 2024 · 75 citations)
Add SMPL-X initialization and pose conditioningText-driven 3D human generation · body structure coherenceHelpedHumanGaussian: Text-Driven 3D Human Generation with Gaussian Splatting (CVPR 2024 · 60 citations)
use implicit Gaussian attribute fields instead of directly optimizing Gaussian attributestext-prompted avatar generation · avatar quality and optimization stabilityHelpedGAvatar: Animatable 3D Gaussian Avatars with Implicit Mesh Learning (CVPR 2024 · 40 citations)
Use 5 instead of 30+ camera viewsMANUS-Grasps, Subjects 1–3 · mIoUSubject1: 0.147 vs 0.206; Subject2: 0.140 vs 0.152; Subject3: 0.214 vs 0.275HurtMANUS: Markerless Grasp Capture using Articulated 3D Gaussians (CVPR 2024 · 23 citations)
use latent space instead of pixel space2K2K evaluation · PSNR23.125 -> 24.293HelpedHumanSplat: Generalizable Single-Image Human Gaussian Splatting with Structure Priors (NeurIPS 2024 · 19 citations)
add progressive optimization rather than jointly optimizing all parametersPARIS dataset, Foldchair start state · PSNR31.49 -> 41.68HelpedArticulatedGS: Self-supervised Digital Twin Modeling of Articulated Objects using 3D Gaussian Splatting (CVPR 2025 · 16 citations)
Use the clothed SMPLX++ template instead of SMPLX directlyFull-body talking · PSNR28.47 -> 33.29HelpedTaoAvatar: Real-Time Lifelike Full-Body Talking Avatars for Augmented Reality via 3D Gaussian Splatting (CVPR 2025 · 12 citations)
add diffuse shadingNovel pose (Synthetic Sequence) · PSNR21.22 -> 22.30HelpedGaussian Shadow Casting for Neural Characters (CVPR 2024 · 12 citations)
add multi-view diffusion priorsNeRSemble novel views · LPIPS0.207 -> 0.118HelpedGAF: Gaussian Avatar Reconstruction from Monocular Videos via Multi-view Diffusion (CVPR 2025 · 9 citations)
add the warm-up stage using pseudo GaussiansINSTA · PSNR8.97 -> 30.52HelpedGraphAvatar: Compact Head Avatars with GNN-Generated 3D Gaussians (AAAI 2025 · 7 citations)

What reviewers will expect

Datasets and baselines named in the 20 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.

Datasets

  • NeRSemble · 3 of 20 papers
  • PartNet-Mobility · 3 of 20 papers
  • Arctic · 2 of 20 papers
  • THuman2.0 · 2 of 20 papers
  • MultiScan · 2 of 20 papers
  • Rat7M · 2 of 20 papers

Baselines

  • Ditto · 4 of 20 papers
  • PARIS · 4 of 20 papers
  • FlashAvatar · 3 of 20 papers
  • GaussianAvatars · 3 of 20 papers
  • INSTA · 2 of 20 papers
  • NeRFBlendShape · 2 of 20 papers
  • ROME · 2 of 20 papers
  • GPS-Gaussian · 2 of 20 papers

Recent preprints

Loading recent arXiv preprints…

    Who publishes on it

    Institutions on the most papers of this approach since 2023 (of 42 papers; from OpenAlex and the papers' affiliations).

    Use it

    Download BibTeX of 20 key papersOpen the topic in the interactive map

    Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.

    Same keyword cluster

    Elsewhere in the topic

    Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).

    Get the next scime Dispatch and a monthly field update

    Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy