SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.03954 · CVPR · 2021

De-rendering the World's Revolutionary Artefacts

Jiajun Wu, Ameesh Makadia, Noah Snavely, Angjoo Kanazawa, Shangzhe Wu, Richard Tucker

arXiv · PDF · Open in the Atlas

Code that ran

10 of 12 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 2 have no contract verdict; 6 failures of our instrument; 2 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/2104.03954.json · badge: /api/ran/2104.03954.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
elliottwu/sorderender canonical (classifier label)
classifier basis: census
10 ran · 2 no recorded run
FunctionStatusWhere it lives
gamma Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/eval/eval_syn_vase.py
code served (permissive licence) · get_code("22c0675924a1c807")
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)

# The function gamma takes an img parameter and performs:
# 1. img.clamp(min=EPS) - clamps tensor values to minimum EPS
# 2. ** GAMMA - raises to power GAMMA
# So img must be a torch.Tensor

img = torch.randn(2, 4)

ARGS = {
    "img": img
}
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
get_rotation_matrix Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/derender/rendering.py
code served (permissive licence) · get_code("ad4c1cc0caa8375d")
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, 3, 3) 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)

# The function expects rxyz to be a tensor with shape (batch_size, 3)
# where the 3 values are rx, ry, rz (rotation angles in radians)
rxyz = torch.randn(2, 3)

ARGS = {
    "rxyz": rxyz
}
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, 3, 3) — shape (2, 3, 3) 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
is_image_file Ran
ran on a synthesized input
elliottwu/sorderender/derender/dataloaders.py
code served (permissive licence) · get_code("521cedf170be4833")
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 bool, 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

# The function is_image_file takes a filename and checks if it ends with
# one of the IMG_EXTENSIONS. It needs a single string parameter.

ARGS = {
    'filename': 'test_image.jpg'
}
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 bool, 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
load_imgs Ran
ran, honoured the contract we drafted
elliottwu/sorderender/render_animation.py
code served (permissive licence) · get_code("fc8963e6961c9146")
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, 3) 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 cv2
import torch
import numpy as np
import tempfile
import os

torch.manual_seed(0)

# Create temporary image files for testing
temp_dir = tempfile.mkdtemp()

# Create 2 small dummy images (4x3 BGR images)
img1 = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)
img2 = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)

# Write images to temporary files
img1_path = os.path.join(temp_dir, "img1.png")
img2_path = os.path.join(temp_dir, "img2.png")

cv2.imwrite(img1_path, img1)
cv2.imwrite(img2_path, img2)

flist = [img1_path, img2_path]

ARGS = {
    "flist": flist
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy, 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, 3) — shape (2, 3, 4, 3) 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
load_txts Ran
ran, honoured the contract we drafted
elliottwu/sorderender/render_animation.py
code served (permissive licence) · get_code("15f1f655fa311bd8")
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, 3) 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 numpy as np
import torch
import tempfile
import os

torch.manual_seed(0)

# Create temporary CSV files with small numeric data
temp_dir = tempfile.mkdtemp()
flist = []

for i in range(2):
    filepath = os.path.join(temp_dir, f"data_{i}.csv")
    # Create a small 4x3 array (batch size 2, features distinct from dims)
    data = np.random.randn(4, 3)
    np.savetxt(filepath, data, delimiter=',')
    flist.append(filepath)

ARGS = {
    "flist": flist
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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, 3) — shape (2, 4, 3) 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
load_yaml Ran
ran on a synthesized input
elliottwu/sorderender/derender/utils.py
code served (permissive licence) · get_code("c9d96366dfea8bc9")
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 dict. 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 tempfile
import os
import yaml

# Create a temporary YAML file with valid content
temp_dir = tempfile.mkdtemp()
yaml_path = os.path.join(temp_dir, "test_config.yaml")

# Write a simple valid YAML file
config_data = {
    "model": "test_model",
    "batch_size": 32,
    "learning_rate": 0.001,
    "layers": [64, 128, 256]
}

with open(yaml_path, 'w') as f:
    yaml.dump(config_data, f)

ARGS = {
    "path": yaml_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
yaml — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
make_dataset Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/derender/dataloaders.py
code served (permissive licence) · get_code("641352b2fce0b79c")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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
import shutil

# Create a temporary directory with some test image files
test_dir = tempfile.mkdtemp()

# Create some image files with various extensions
image_files = [
    'image1.jpg',
    'image2.png',
    'image3.jpeg',
    'image4.bmp',
    'image5.tiff',
    'not_an_image.txt',  # This should be ignored
]

for fname in image_files:
    fpath = os.path.join(test_dir, fname)
    # Create empty files
    with open(fpath, 'w') as f:
        f.write('')

# Also create a subdirectory with more images
subdir = os.path.join(test_dir, 'subdir')
os.makedirs(subdir, exist_ok=True)
subimage_files = ['subimage1.webp', 'subimage2.ppm']
for fname in subimage_files:
    fpath = os.path.join(subdir, fname)
    with open(fpath, 'w') as f:
        f.write('')

ARGS = {
    'dir': test_dir
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
rotate_pts Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/derender/rendering.py
code served (permissive licence) · get_code("08ef352332219435")
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, 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
import torch.nn as nn

torch.manual_seed(0)

# Create sample tensors
# pts: shape (batch_size, num_points, input_dim)
# rotmat: shape (batch_size, output_dim, input_dim)

pts = torch.randn(2, 4, 3)  # batch=2, num_points=4, input_dim=3
rotmat = torch.randn(2, 5, 3)  # batch=2, output_dim=5, input_dim=3

ARGS = {
    "pts": pts,
    "rotmat": rotmat
}
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, 5) — shape (2, 4, 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
setup_runtime Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/derender/utils.py
code served (permissive licence) · get_code("dd5d55f35cba2c2c")
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 dict. 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 random
import numpy as np
import torch
import yaml
import tempfile

# Create a temporary YAML config file for testing
temp_config = tempfile.NamedTemporaryFile(mode='w', suffix='.yaml', delete=False)
yaml.dump({'model': 'test_model', 'batch_size': 32}, temp_config)
temp_config.close()
config_path = temp_config.name

# Create a simple args object
class Args:
    def __init__(self):
        self.gpu = None  # CPU only
        self.seed = 42
        self.config = config_path
        self.num_workers = 4

args_obj = Args()

ARGS = {
    'args': args_obj
}

# Cleanup is handled by the test framework, but we can note the temp file path
# The actual test will use this ARGS dict to call setup_runtime
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, yaml — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
translate_pts Ran
ran; our contract draft was wrong, not the code
elliottwu/sorderender/derender/rendering.py
code served (permissive licence) · get_code("fc4db8f56b5c70e9")
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, 3) 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)

# pts: tensor of points, shape (batch, num_points, dims)
# txyz: tensor of translations, shape (batch, dims)
# The function does: pts + txyz.unsqueeze(1)
# unsqueeze(1) expands txyz from (batch, dims) to (batch, 1, dims)
# Then broadcasting adds it to pts

pts = torch.randn(2, 4, 3)  # batch=2, num_points=4, dims=3
txyz = torch.randn(2, 3)   # batch=2, dims=3

ARGS = {
    "pts": pts,
    "txyz": txyz,
}
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, 3) — shape (2, 4, 3) 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
dump_yaml Not yet run
unverified (harvested, no recorded run)
elliottwu/sorderender/derender/utils.py
code served (permissive licence) · get_code("8e5f27fa4d58f14b")
How it ran

no invocation recorded.

get_data_loaders Not yet run
unverified (harvested, no recorded run)
elliottwu/sorderender/derender/dataloaders.py
code served (permissive licence) · get_code("afcb544a47dec12a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 2 later) and 9 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
load_yaml 4 other papers
2 dated earlier, 2 later
1911.11130 (2019-11)
Unsupervised Learning of Probably Symmetric Deformable 3D Objects from Images in the Wild
2011.00844 (2020-11)
Do 2D GANs Know 3D Shape? Unsupervised 3D shape reconstruction from 2D Image GANs
2109.02288 (2021-09)
Toward Realistic Single-View 3D Object Reconstruction with Unsupervised Learning from Multiple Images
and 1 more not named here
is_image_file 3 other papers
1 dated earlier, 2 later
1911.11130 (2019-11)
Unsupervised Learning of Probably Symmetric Deformable 3D Objects from Images in the Wild
2109.02288 (2021-09)
Toward Realistic Single-View 3D Object Reconstruction with Unsupervised Learning from Multiple Images
2201.02279 (2022-01)
De-rendering 3D Objects in the Wild
make_dataset 1 other paper
1 dated earlier
1911.11130 (2019-11)
Unsupervised Learning of Probably Symmetric Deformable 3D Objects from Images 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

Albedo Diffuse Specular Material Env. map Normal Novel view Single image collection Relight Figure 1: De-rendering from single images. From only a real single-view image collection of "revolutionary" (i.e., solid of revolution) artefacts with known silhouettes as training data (left), our framework learns to de-render a single image into shape, albedo and complex lighting and material components, suitable for applications such as novel-view synthesis and relighting (right).

For agents

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

get_harvested_code_for_paper("2104.03954")
get_code_for_paper("2104.03954")
have("2104.03954")

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

curl https://syntology.ai/api/ran/2104.03954.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.03954.svg)](https://syntology.ai/paper/2104.03954)

Connect an agent — have() is free.