SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2103.17185 · CVPR · 2021

Rethinking Style Transfer: From Pixels to Parameterized Brushstrokes

Björn Ommer, Dmytro Kotovenko, Matthias Wright, Arthur Heimbrecht

arXiv · PDF · Open in the Atlas

Code that ran

2 of 16 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 1 failure of our instrument; 14 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/2103.17185.json · badge: /api/ran/2103.17185.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
CompVis/brushstroke-parameterized-style-transfer marked official in Papers with Code
Papers with Code archive
2 ran · 14 no recorded run
FunctionStatusWhere it lives
VGG Ran
ran on a synthesized input · constructed an object; no result was computed
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("6f0ac111ae4ae336")
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 an instance of its own class, VGG: the run constructed the object and did not call it, so no result was computed. 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 pickle
import tempfile
import os
import numpy as np

# Create a temporary checkpoint file with the required parameter structure
param_dict = {
    'block1': {
        'conv1': {
            'weight': np.random.randn(3, 3, 3, 64).astype(np.float32),
            'bias': np.random.randn(64).astype(np.float32)
        },
        'conv2': {
            'weight': np.random.randn(3, 3, 64, 64).astype(np.float32),
            'bias': np.random.randn(64).astype(np.float32)
        }
    },
    'block2': {
        'conv1': {
            'weight': np.random.randn(3, 3, 64, 128).astype(np.float32),
            'bias': np.random.randn(128).astype(np.float32)
        },
        'conv2': {
            'weight': np.random.randn(3, 3, 128, 128).astype(np.float32),
            'bias': np.random.randn(128).astype(np.float32)
        }
    },
    'block3': {
        'conv1': {
            'weight': np.random.randn(3, 3, 128, 256).astype(np.float32),
            'bias': np.random.randn(256).astype(np.float32)
        },
        'conv2': {
            'weight': np.random.randn(3, 3, 256, 256).astype(np.float32),
            'bias': np.random.randn(256).astype(np.float32)
        },
        'conv3': {
            'weight': np.random.randn(3, 3, 256, 256).astype(np.float32),
            'bias': np.random.randn(256).astype(np.float32)
        }
    },
    'block4': {
        'conv1': {
            'weight': np.random.randn(3, 3, 256, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        },
        'conv2': {
            'weight': np.random.randn(3, 3, 512, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        },
        'conv3': {
            'weight': np.random.randn(3, 3, 512, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        }
    },
    'block5': {
        'conv1': {
            'weight': np.random.randn(3, 3, 512, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        },
        'conv2': {
            'weight': np.random.randn(3, 3, 512, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        },
        'conv3': {
            'weight': np.random.randn(3, 3, 512, 512).astype(np.float32),
            'bias': np.random.randn(512).astype(np.float32)
        }
    }
}

# Create a temporary file to store the pickle
temp_fd, ckpt_path = tempfile.mkstemp(suffix='.pkl')
os.close(temp_fd)
with open(ckpt_path, 'wb') as f:
    pickle.dump(param_dict, f)

ARGS = {
    'ckpt_path': ckpt_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type VGG — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a VGG 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
initialize_brushstrokes Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("fadf566cf66fcbbf")
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. 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 numpy as np

# Create a small content image (4x5x3 RGB image)
content_img = np.random.rand(4, 5, 3).astype(np.float32) * 255

num_strokes = 3
canvas_height = 64
canvas_width = 80
sec_scale = 0.001
width_scale = 1.0
init = 'sp'

ARGS = {
    'content_img': content_img,
    'num_strokes': num_strokes,
    'canvas_height': canvas_height,
    'canvas_width': canvas_width,
    'sec_scale': sec_scale,
    'width_scale': width_scale,
    'init': init
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, skimage — 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
BrushstrokeOptimizer Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("738f825697601e72")
How it ran

no invocation recorded.

clusters_to_strokes Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("aa723ef48d78c84a")
How it ran

no invocation recorded.

content_loss Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("11b86db023656ede")
How it ran

no invocation recorded.

curviture_loss Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("8fb99ce0ede1ac50")
How it ran

no invocation recorded.

download_weights Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("d0a44dffd791518a")
How it ran

no invocation recorded.

draw_projection_loss Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("eaaca87d3d35ace6")
How it ran

no invocation recorded.

get_gram_matrices Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("85360c429d1aca56")
How it ran

no invocation recorded.

get_nn_idxs Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("4decbae4c04176d8")
How it ran

no invocation recorded.

norm Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("8fb69f821ee3a813")
How it ran

no invocation recorded.

preprocess_img Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("00a501f8bebc184e")
How it ran

no invocation recorded.

renderer Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("5457d2c906d82b7e")
How it ran

no invocation recorded.

sample_quadratic_bezier_curve Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("e0a52f21c9ffce5a")
How it ran

no invocation recorded.

style_loss Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("3bc3bc04c01ea4bb")
How it ran

no invocation recorded.

total_variation_loss Not yet run
unverified (harvested, no recorded run)
CompVis/brushstroke-parameterized-style-transfer/model.py
code served (permissive licence) · get_code("0ff29180d1bf839d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 16 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.

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

There have been many successful implementations of neural style transfer in recent years. In most of these works, the stylization process is confined to the pixel domain. However, we argue that this representation is unnatural because paintings usually consist of brushstrokes rather than pixels. We propose a method to stylize images by optimizing parameterized brushstrokes instead of pixels and further introduce a simple differentiable rendering mechanism. Our approach significantly improves visual quality and enables additional control over the stylization process such as controlling the flow of brushstrokes through user input. We provide qualitative and quantitative evaluations that show the efficacy of the proposed parameterized representation. Code is available at https://github. com / CompVis / brushstroke -parameterizedstyle-transfer.

For agents

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

get_harvested_code_for_paper("2103.17185")
get_code_for_paper("2103.17185")
have("2103.17185")

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

curl https://syntology.ai/api/ran/2103.17185.json

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

[![Syntology run record](https://syntology.ai/api/ran/2103.17185.svg)](https://syntology.ai/paper/2103.17185)

Connect an agent — have() is free.