SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2011.12438 · NeurIPS · 2020

Continuous Surface Embeddings

Andrea Vedaldi, David Novotny, Natalia Neverova, Vasil Khalidov, Marc Szafraniec, Patrick Labatut

arXiv · PDF · Open in the Atlas

Code that ran

5 of 6 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; 1 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 3 failures of our instrument; 1 no recorded run), measured 2026-11-21; 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/2011.12438.json · badge: /api/ran/2011.12438.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
zhongshsh/4d-animal community
Papers with Code archive
5 ran · 1 no recorded run
FunctionStatusWhere it lives
CseProcessing Ran
ran on a synthesized input · constructed an object; no result was computed
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("b597401c9c6ca9b2")
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, CseProcessing: 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 os
import torch
import pickle as pk
import numpy as np
import tempfile
import shutil

torch.manual_seed(0)

# Create a temporary directory structure for the CSE processing
temp_dir = tempfile.mkdtemp()
cse_data_root_path = temp_dir
lbo_data_folder = temp_dir

# Create subdirectories
os.makedirs(os.path.join(lbo_data_folder, 'lbos'), exist_ok=True)
os.makedirs(os.path.join(lbo_data_folder, 'cse'), exist_ok=True)

# Create mock CSE preprocessing data file
sequence_index = "0001_001_0001"
frame_limit = 10
cse_predictions_path = os.path.join(cse_data_root_path, f'{sequence_index}_{frame_limit}_cse_predictions.pk')

mock_cse_data = {
    'scores': [0.95, 0.92, 0.88, 0.91, 0.89, 0.93, 0.90, 0.87, 0.94, 0.92],
    'embeddings': np.random.randn(10, 256).astype(np.float32),
    'uv_coords': np.random.randn(10, 100, 2).astype(np.float32),
}

with open(cse_predictions_path, 'wb') as f:
    pk.dump(mock_cse_data, f)

# Create mock LBO data files
mock_lbo_cse = {
    'lbo_basis': np.random.randn(500, 256).astype(np.float32),
    'lbo_areas': np.ones(500).astype(np.float32) / 500,
}

lbo_cse_path = os.path.join(lbo_data_folder, 'lbos', 'lbo_cse.pk')
with open(lbo_cse_path, 'wb') as f:
    pk.dump(mock_lbo_cse, f)

mock_lbo_smal = {
    'lbo_basis': np.random.randn(3000, 256).astype(np.float32),
    'lbo_areas': np.ones(3000).astype(np.float32) / 3000,
}

lbo_smal_path = os.path.join(lbo_data_folder, 'lbos', 'lbo_smal.pk')
with open(lbo_smal_path, 'wb') as f:
    pk.dump(mock_lbo_smal, f)

ARGS = {
    'cse_data_root_path': cse_data_root_path,
    'sequence_index': sequence_index,
    'densepose_version': 'V2_I2M',
    'lbo_data_folder': lbo_data_folder,
    'frame_limit': frame_limit,
    'category': 'dog',
}
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 CseProcessing — 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 CseProcessing 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_lbo_from_name Ran
ran; our contract draft was wrong, not the code
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("3a59c264202f943d")
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 os
import pickle as pk
import numpy as np
import tempfile

# Create a temporary directory for test files
temp_dir = tempfile.mkdtemp()

# Create test LBO data
lbo_data = {
    'lbo_basis': np.random.randn(100, 256).astype(np.float32),
    'lbo_areas': np.random.randn(100, 100).astype(np.float32)
}

# Save test pickle file
mesh_name = 'smal'
lbo_filepath = os.path.join(temp_dir, f'lbo_{mesh_name}.pk')
with open(lbo_filepath, 'wb') as f:
    pk.dump(lbo_data, f)

ARGS = {
    'mesh_name': 'smal',
    'query_folder': temp_dir
}
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 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
get_mapping_from_name Ran
ran, honoured the contract we drafted
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("d730689f8c32a54f")
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 ndarray, ndim 2, dtype family float, shape (128, 128) 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 os
import pickle as pk
import numpy as np
import tempfile
import shutil

# Create a temporary directory for the query folder
temp_dir = tempfile.mkdtemp()

# Create a valid pickle file with the expected structure
mesh_source_name = "mesh_a"
mesh_destination_name = "mesh_b"
mapping_filepath = os.path.join(temp_dir, f'lbo_{mesh_source_name}_to_{mesh_destination_name}.pk')

# Create sample data with the expected 'C' key containing a (128, 128) float32 array
sample_data = {
    'C': np.random.randn(128, 128).astype(np.float32)
}

# Write the pickle file
with open(mapping_filepath, 'wb') as f:
    pk.dump(sample_data, f)

ARGS = {
    'mesh_source_name': mesh_source_name,
    'mesh_destination_name': mesh_destination_name,
    'query_folder': temp_dir
}
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 ndarray, ndim 2, dtype family float, shape (128, 128) — shape (128, 128) 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_numpy Ran
ran; our contract draft was wrong, not the code
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("13134f47b0fecb82")
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 ndarray, 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 numpy as np

# Create a small numpy array with shape [N, D] = [2, 4]
# N=2 embedding vectors, each with D=4 dimensions
embeddings = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0]
], dtype=np.float32)

ARGS = {
    "embeddings": embeddings,
    "dim": 1,
    "epsilon": 1e-06
}
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 ndarray, 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
resize_torch Ran
ran; our contract draft was wrong, not the code
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("1567009b1e58288f")
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, 6, 7, 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.functional as F

torch.manual_seed(0)

# Test case 1: 4D tensor (BATCH, H, W, CHANNEL)
X_4d = torch.randn(2, 4, 3, 5, dtype=torch.float32)

# Test case 2: 3D tensor (H, W, CHANNEL)
X_3d = torch.randn(4, 3, 5, dtype=torch.float32)

# Test case 3: List of tensors
X_list = [torch.randn(4, 3, 5, dtype=torch.float32), torch.randn(2, 4, 3, 5, dtype=torch.float32)]

# Target size
size_tuple = (6, 7)

# Mode for interpolation
mode_str = 'bilinear'

# Create ARGS dict with all parameter combinations to test
ARGS = {
    'X': X_4d,
    'size': size_tuple,
    'mode': mode_str
}
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, 6, 7, 5) — shape (2, 6, 7, 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
map_X_from_source_to_destination Not yet run
unverified (harvested, no recorded run)
zhongshsh/4d-animal/cse_embedding/cse_embedder.py
code served (permissive licence) · get_code("ec94e6d72be4aaff")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 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 6 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 2026-11-21.

Repositories linked to this paper

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

Abstract

In this work, we focus on the task of learning and representing dense correspondences in deformable object categories. While this problem has been considered before, solutions so far have been rather ad-hoc for specific object types (i.e., humans), often with significant manual work involved. However, scaling the geometry understanding to all objects in nature requires more automated approaches that can also express correspondences between related, but geometrically different objects. To this end, we propose a new, learnable image-based representation of dense correspondences. Our model predicts, for each pixel in a 2D image, an embedding vector of the corresponding vertex in the object mesh, therefore establishing dense correspondences between image pixels and 3D object geometry. We demonstrate that the proposed approach performs on par or better than the state-ofthe-art methods for dense pose estimation for humans, while being conceptually simpler. We also collect a new in-the-wild dataset of dense correspondences for animal classes and demonstrate that our framework scales naturally to the new deformable object categories. 34th Conference on Neural Information Processing Systems (NeurIPS 2020),

For agents

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

get_harvested_code_for_paper("2011.12438")
get_code_for_paper("2011.12438")
have("2011.12438")

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

curl https://syntology.ai/api/ran/2011.12438.json

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

[![Syntology run record](https://syntology.ai/api/ran/2011.12438.svg)](https://syntology.ai/paper/2011.12438)

Connect an agent — have() is free.