SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.06149 · ECCV · 2024

HeadStudio: Text to Animatable Head Avatars with 3D Gaussian Splatting

Yi Yang, Zongxin Yang, Hehe Fan, Fan Ma, Zhenglin Zhou

arXiv · PDF · Open in the Atlas

Code that ran

5 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 3 failures of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2402.06149.json · badge: /api/ran/2402.06149.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
ZhenglinZhou/HeadStudio canonical (classifier label)
classifier basis: census
5 ran · 3 no recorded run
FunctionStatusWhere it lives
gaussian Ran
ran, honoured the contract we drafted
ZhenglinZhou/HeadStudio/gaussiansplatting/utils/loss_utils.py
code served (permissive licence) · get_code("c56b7ef16f309a45")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 1, dtype family float, shape (5,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from math import exp

torch.manual_seed(0)

ARGS = {
    "window_size": 5,
    "sigma": 1.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (5,) — shape (5,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_c2w Ran
ran; our contract draft was wrong, not the code
ZhenglinZhou/HeadStudio/animation.py
code served (permissive licence) · get_code("dcc7d8db6ac85aa5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 4, 4) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import math

torch.manual_seed(0)

# Create small tensors for the function parameters
batch_size = 2

# dist: distance values, shape [B]
dist = torch.tensor([1.0, 2.0], dtype=torch.float32)

# elev: elevation angles in degrees, shape [B]
elev = torch.tensor([30.0, 45.0], dtype=torch.float32)

# azim: azimuth angles in degrees, shape [B]
azim = torch.tensor([0.0, 90.0], dtype=torch.float32)

# device: CPU device
device = torch.device('cpu')

ARGS = {
    'dist': dist,
    'elev': elev,
    'azim': azim,
    'device': device
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 4) — shape (2, 4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
l1_loss Ran
ran; our contract draft was wrong, not the code
ZhenglinZhou/HeadStudio/gaussiansplatting/utils/loss_utils.py
code served (permissive licence) · get_code("ac0e42d6fbcfbbe6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small tensors for network output and ground truth
# Both should have the same shape for the subtraction and mean operations
network_output = torch.randn(2, 4)  # batch size 2, 4 dimensions
gt = torch.randn(2, 4)  # batch size 2, 4 dimensions

ARGS = {
    "network_output": network_output,
    "gt": gt,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
l2_loss Ran
ran, honoured the contract we drafted
ZhenglinZhou/HeadStudio/gaussiansplatting/utils/loss_utils.py
code served (permissive licence) · get_code("8c3b0f873ba11813")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create small tensors for network output and ground truth
# Both should have the same shape for the loss computation
# Using batch size 2, feature dimension 4
network_output = torch.randn(2, 4)
gt = torch.randn(2, 4)

ARGS = {
    "network_output": network_output,
    "gt": gt,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
readImages Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ZhenglinZhou/HeadStudio/gaussiansplatting/metrics.py
code served (permissive licence) · get_code("cdd00787894554b5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
from pathlib import Path
from PIL import Image
import torch
import numpy as np

torch.manual_seed(0)

# Create temporary directories with dummy image files
temp_renders_dir = tempfile.mkdtemp()
temp_gt_dir = tempfile.mkdtemp()

renders_dir = Path(temp_renders_dir)
gt_dir = Path(temp_gt_dir)

# Create 2 small dummy images (4x3 RGB)
for i in range(2):
    # Create render image
    render_array = np.random.randint(0, 256, (3, 4, 3), dtype=np.uint8)
    render_img = Image.fromarray(render_array.transpose(1, 2, 0), mode='RGB')
    render_img.save(renders_dir / f"render_{i}.png")
    
    # Create ground truth image
    gt_array = np.random.randint(0, 256, (3, 4, 3), dtype=np.uint8)
    gt_img = Image.fromarray(gt_array.transpose(1, 2, 0), mode='RGB')
    gt_img.save(gt_dir / f"render_{i}.png")

ARGS = {
    "renders_dir": renders_dir,
    "gt_dir": gt_dir,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
batch_gs_render Not yet run
unverified (harvested, no recorded run)
ZhenglinZhou/HeadStudio/animation.py
code served (permissive licence) · get_code("2264ffba1d2ef541")
How it ran

no invocation recorded.

batch_gs_render Not yet run
unverified (harvested, no recorded run)
ZhenglinZhou/HeadStudio/animation_TalkSHOW.py
code served (permissive licence) · get_code("40c6045859b09953")
How it ran

no invocation recorded.

batch_mesh_render Not yet run
unverified (harvested, no recorded run)
ZhenglinZhou/HeadStudio/animation.py
code served (permissive licence) · get_code("644f4722f144360f")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 4 the same month, 4 later) and 4 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
gaussian 160 other papers
69 dated earlier, 3 the same month, 87 later, 1 same year, order not determinable · 7 only as an assembly brick
1812.11941 (2018-12)
High Quality Monocular Depth Estimation via Transfer Learning
1904.01538 (2019-04)
Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset
1907.11628 (2019-07)
Unsupervised Learning for Optical Flow Estimation Using Pyramid Convolution LSTM
and 157 more not named here
l1_loss 93 other papers
20 dated earlier, 3 the same month, 70 later
2308.04079 (2023-08)
3D Gaussian Splatting for Real-Time Radiance Field Rendering
2309.13101 (2023-09)
Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction
2310.08528 (2023-10)
4D Gaussian Splatting for Real-Time Dynamic Scene Rendering
and 90 more not named here
l2_loss 90 other papers
21 dated earlier, 3 the same month, 66 later
2308.04079 (2023-08)
3D Gaussian Splatting for Real-Time Radiance Field Rendering
2309.13101 (2023-09)
Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction
2310.08528 (2023-10)
4D Gaussian Splatting for Real-Time Dynamic Scene Rendering
and 87 more not named here
readImages 54 other papers
13 dated earlier, 3 the same month, 38 later
2308.04079 (2023-08)
3D Gaussian Splatting for Real-Time Radiance Field Rendering
2309.13101 (2023-09)
Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction
2310.08528 (2023-10)
4D Gaussian Splatting for Real-Time Dynamic Scene Rendering
and 51 more not named here

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Creating digital avatars from textual prompts has long been a desirable yet challenging task. Despite the promising results achieved with 2D diffusion priors, current methods struggle to create high-quality and consistent animated avatars efficiently. Previous animatable head models like FLAME have difficulty in accurately representing detailed texture and geometry. Additionally, high-quality 3D static representations face challenges in semantically driving with dynamic priors. In this paper, we introduce HeadStudio, a novel framework that utilizes 3D Gaussian splatting to generate realistic and animatable avatars from text prompts. Firstly, we associate 3D Gaussians with animatable head prior model, facilitating semantic animation on high-quality 3D representations. To ensure consistent animation, we further enhance the optimization from initialization, distillation, and regularization to jointly learn the shape, texture, and animation. Extensive experiments demonstrate the efficacy of HeadStudio in generating animatable avatars from textual prompts, exhibiting appealing appearances. The avatars are capable of rendering high-quality real-time (≥ 40 fps) novel views at a resolution of 1024. Moreover, These avatars can be smoothly driven by real-world speech and video. We hope that HeadStudio can enhance digital avatar creation and gain popularity in the community. Code is at: https://github.com/ZhenglinZhou/HeadStudio.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2402.06149")
get_code_for_paper("2402.06149")
have("2402.06149")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2402.06149.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2402.06149.svg)](https://syntology.ai/paper/2402.06149)

Connect an agent — have() is free.