SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.04036 · ECCV · 2022

StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via Pre-trained StyleGAN

Baoyuan Wu, Yong Zhang, Yanbo Fan, Jue Wang, Yujiu Yang, Qingyan Bai, Xuan Wang, Fei Yin, Cun, Mingdeng Cao

arXiv · PDF · Open in the Atlas

Code that ran

4 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; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 2 failures of our instrument; 4 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/2203.04036.json · badge: /api/ran/2203.04036.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
FeiiYin/StyleHEAT canonical (classifier label)
classifier basis not recorded on the link
4 ran · 4 no recorded run
FunctionStatusWhere it lives
apply_imagenet_normalization Ran
ran on a synthesized input
FeiiYin/StyleHEAT/loss/perceptual.py
code served (permissive licence) · get_code("1f603586a8f2d523")
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 4, dtype family float, shape (2, 3, 4, 4) 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 a 4D tensor NxCxHxW with values in [-1, 1]
# N=2 (batch size), C=3 (RGB channels), H=4, W=4 (spatial dimensions)
input_tensor = torch.rand(2, 3, 4, 4) * 2 - 1  # Scale to [-1, 1]

ARGS = {
    "input": input_tensor
}
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 4, dtype family float, shape (2, 3, 4, 4) — shape (2, 3, 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
normalize_activation Ran
ran; our contract draft was wrong, not the code
FeiiYin/StyleHEAT/loss/lpips.py
code served (permissive licence) · get_code("1dab900b2adbe38e")
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 2, dtype family float, shape (2, 4) 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 a small batch of activation values
# Shape: (batch_size=2, features=4)
x = torch.randn(2, 4)

ARGS = {
    "x": x,
    "eps": 1e-10
}
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 2, dtype family float, shape (2, 4) — shape (2, 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
recursive_update Ran
ran on a synthesized input
FeiiYin/StyleHEAT/configs/config.py
code served (permissive licence) · get_code("f928ccb549fdb06b")
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 AttrDict. The image it ran in was 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 collections

class AttrDict(dict):
    """Dict as attribute trick."""

    def __init__(self, *args, **kwargs):
        super(AttrDict, self).__init__(*args, **kwargs)
        self.__dict__ = self
        for key, value in self.__dict__.items():
            if isinstance(value, dict):
                self.__dict__[key] = AttrDict(value)
            elif isinstance(value, (list, tuple)):
                if isinstance(value[0], dict):
                    self.__dict__[key] = [AttrDict(item) for item in value]
                else:
                    self.__dict__[key] = value

    def yaml(self):
        """Convert object to yaml dict and return."""
        yaml_dict = {}
        for key, value in self.__dict__.items():
            if isinstance(value, AttrDict):
                yaml_dict[key] = value.yaml()
            elif isinstance(value, list):
                if isinstance(value[0], AttrDict):
                    new_l = []
                    for item in value:
                        new_l.append(item.yaml())
                    yaml_dict[key] = new_l
                else:
                    yaml_dict[key] = value
            else:
                yaml_dict[key] = value
        return yaml_dict

    def __repr__(self):
        """Print all variables."""
        ret_str = []
        for key, value in self.__dict__.items():
            if isinstance(value, AttrDict):
                ret_str.append('{}:'.format(key))
                child_ret_str = value.__repr__().split('\n')
                for item in child_ret_str:
                    ret_str.append('    ' + item)
            elif isinstance(value, list):
                if isinstance(value[0], AttrDict):
                    ret_str.append('{}:'.format(key))
                    for item in value:
                        child_ret_str = item.__repr__().split('\n')
                        for item in child_ret_str:
                            ret_str.append('    ' + item)
                else:
                    ret_str.append('{}: {}'.format(key, value))
            else:
                ret_str.append('{}: {}'.format(key, value))
        return '\n'.join(ret_str)

def recursive_update(d, u):
    """Recursively update AttrDict d with AttrDict u"""
    for key, value in u.items():
        if isinstance(value, collections.abc.Mapping):
            d.__dict__[key] = recursive_update(d.get(key, AttrDict({})), value)
        elif isinstance(value, (list, tuple)):
            if isinstance(value[0], dict):
                d.__dict__[key] = [AttrDict(item) for item in value]
            else:
                d.__dict__[key] = value
        else:
            d.__dict__[key] = value
    return d

# Setup for testing recursive_update function
d = AttrDict({
    'name': 'base_config',
    'learning_rate': 0.001,
    'nested': {
        'param1': 10,
        'param2': 20
    },
    'list_of_dicts': [
        {'key': 'val1'},
        {'key': 'val2'}
    ]
})

u = AttrDict({
    'learning_rate': 0.01,
    'nested': {
        'param1': 100,
        'param3': 30
    },
    'new_param': 'new_value',
    'list_of_dicts': [
        {'key': 'updated_val1'},
        {'key': 'updated_val2'}
    ]
})

ARGS = {
    'd': d,
    'u': u
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type AttrDict — a AttrDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a AttrDict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not 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
spectral_norm Ran
ran; our contract draft was wrong, not the code
FeiiYin/StyleHEAT/models/styleheat/base_function.py
code served (permissive licence) · get_code("26a6354827d68707")
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 Linear. The image it ran in was 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 torch
import torch.nn as nn
from torch.nn.utils.spectral_norm import spectral_norm as SpectralNorm

torch.manual_seed(0)

# Create a simple neural network module to apply spectral norm to
module = nn.Linear(4, 3)

# Define arguments for the spectral_norm function
ARGS = {
    "module": module,
    "use_spect": True
}
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 Linear — a Linear output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Linear output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not 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
get_network Not yet run
unverified (harvested, no recorded run)
FeiiYin/StyleHEAT/loss/lpips.py
code served (permissive licence) · get_code("c2b838a0b10034c3")
How it ran

no invocation recorded.

get_state_dict Not yet run
unverified (harvested, no recorded run)
FeiiYin/StyleHEAT/loss/lpips.py
pointer only (licence: MIT); the stored copy get_code serves, from instantx-research/instantstyle-plus, is pointer only (licence: NONE) · get_code("17a0ef3b8319a5bc")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from instantx-research/instantstyle-plus, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

rgetattr Not yet run
unverified (harvested, no recorded run)
FeiiYin/StyleHEAT/configs/config.py
code served (permissive licence) · get_code("f6c7280f75271482")
How it ran

no invocation recorded.

rsetattr Not yet run
unverified (harvested, no recorded run)
FeiiYin/StyleHEAT/configs/config.py
code served (permissive licence) · get_code("31aa2f35339fe269")
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, 6 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 6 later) and 2 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
get_network 60 other papers
60 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 57 more not named here
normalize_activation 53 other papers
53 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
2310.10642 (2023-10)
Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting
and 50 more not named here
get_state_dict 52 other papers
52 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
2310.10642 (2023-10)
Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting
and 49 more not named here
rgetattr 13 other papers
13 later · 1 only as an assembly brick
2209.02952 (2022-09)
BiFuse++: Self-supervised and Efficient Bi-projection Fusion for 360 Depth Estimation
2211.05778 (2022-11)
InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions
2212.04501 (2022-12)
Learning Video Representations from Large Language Models
and 10 more not named here
rsetattr 8 other papers
8 later
2212.04501 (2022-12)
Learning Video Representations from Large Language Models
2305.13417 (2023-05)
VISIT: Visualizing and Interpreting the Semantic Information Flow of Transformers
2311.12871 (2023-11)
An Embodied Generalist Agent in 3D World
and 5 more not named here
spectral_norm 1 other paper
1 later
2211.14758 (2022-11)
VideoReTalking: Audio-based Lip Synchronization for Talking Head Video Editing In the Wild

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

One-shot talking face generation aims at synthesizing a highquality talking face video from an arbitrary portrait image, driven by a video or an audio segment. In this work, we provide a solution from a novel perspective that differs from existing frameworks. We first investigate the latent feature space of a pre-trained StyleGAN and discover some excellent spatial transformation properties. Upon the observation, we propose a novel unified framework based on a pre-trained StyleGAN that enables a set of powerful functionalities, i.e., high-resolution video generation, disentangled control by driving video or audio, and flexible face editing. Our framework elevates the resolution of the synthesized talking face to 1024×1024 for the first time, even though the training dataset has a lower resolution. Moreover, our framework allows two types of facial editing, i.e., global editing via GAN inversion and intuitive editing via 3D morphable models. Comprehensive experiments show superior video quality and flexible controllability over state-of-the-art methods.

For agents

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

get_harvested_code_for_paper("2203.04036")
get_code_for_paper("2203.04036")
have("2203.04036")

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

curl https://syntology.ai/api/ran/2203.04036.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.04036.svg)](https://syntology.ai/paper/2203.04036)

Connect an agent — have() is free.