SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1901.04111 · 2019

Fast and Robust Multi-Person 3D Pose Estimation from Multiple Views

arXiv · PDF · Open in the Atlas

Code that ran

15 of 19 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, 1 violated it and 13 have no contract verdict; 1 failure 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/1901.04111.json · badge: /api/ran/1901.04111.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
zju3dv/mvpose canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
perfanalytics/pose2sim community (classifier label)
classifier basis not recorded on the link
13 ran · 4 no recorded run
FunctionStatusWhere it lives
determine_level Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/Pose2Sim.py
code served (permissive licence) · get_code("5968d2aed9a18f44")
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 int, ndim 0, dtype family int, 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 os
import tempfile
import shutil

# Create a temporary directory structure with Config.toml files
temp_root = tempfile.mkdtemp()

# Create a trial folder structure with Config.toml at depth 3
trial_dir = os.path.join(temp_root, "experiment", "trial_1")
os.makedirs(trial_dir, exist_ok=True)
with open(os.path.join(trial_dir, "Config.toml"), "w") as f:
    f.write("[config]\nkey = 'value'\n")

# Create another trial at the same depth
trial_dir_2 = os.path.join(temp_root, "experiment", "trial_2")
os.makedirs(trial_dir_2, exist_ok=True)
with open(os.path.join(trial_dir_2, "Config.toml"), "w") as f:
    f.write("[config]\nkey = 'value'\n")

# Create a Config.toml at the root level (depth 1)
with open(os.path.join(temp_root, "Config.toml"), "w") as f:
    f.write("[config]\nkey = 'value'\n")

ARGS = {
    "config_dir": temp_root
}
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 int, ndim 0, dtype family int, 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
get_markers_path Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/kinematics.py
code served (permissive licence) · get_code("51884303c280eb3f")
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 PosixPath. 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.

from pathlib import Path
import tempfile
import os

# Create a temporary directory with marker XML files
temp_dir = tempfile.TemporaryDirectory()
osim_setup_dir = Path(temp_dir.name)

# Create marker files that the function might look for
marker_files = [
    'Markers_Halpe26.xml',
    'Markers_Halpe68_136.xml',
    'Markers_Coco133.xml',
    'Markers_halpe26lower.xml',
]

for marker_file in marker_files:
    (osim_setup_dir / marker_file).touch()

pose_model = 'halpe26_lower'

ARGS = {
    'pose_model': pose_model,
    'osim_setup_dir': osim_setup_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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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_max_workers Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/common.py
code served (permissive licence) · get_code("1d2337c9c4d8a8f2")
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 int, ndim 0, dtype family int, shape () 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 os
import torch

torch.manual_seed(0)

ARGS = {
    'device': 'cpu'
}
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 int, ndim 0, dtype family int, 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
get_model_path Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/kinematics.py
code served (permissive licence) · get_code("afdc19604ea84caf")
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 PosixPath. 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.

from pathlib import Path

use_simple_model = True
osim_setup_dir = Path("/tmp/osim_setup")

ARGS = {
    "use_simple_model": use_simple_model,
    "osim_setup_dir": osim_setup_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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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_scaling_setup Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/kinematics.py
code served (permissive licence) · get_code("9f5e9f9adf429fa2")
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 PosixPath. 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.

from pathlib import Path
import tempfile
import os

# Create a temporary directory to simulate the osim_setup_dir
temp_dir = tempfile.TemporaryDirectory()
osim_setup_dir = Path(temp_dir.name)

# Create some dummy scaling setup files to match the expected naming convention
scaling_files = [
    'Scaling_Setup_Pose2Sim_Halpe26.xml',
    'Scaling_Setup_Pose2Sim_Halpe68_136.xml',
    'Scaling_Setup_Pose2Sim_Coco133.xml',
]

for file in scaling_files:
    (osim_setup_dir / file).touch()

# Define the arguments
ARGS = {
    'pose_model': 'halpe_26_lower',
    'osim_setup_dir': osim_setup_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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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_smoothing_spline_intermediate_arrays Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/filtering.py
code served (permissive licence) · get_code("414564c9c3cdbac8")
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 small, valid input arrays for the function
# x must be ascending, 1D array with at least 5 elements
x = np.array([0.0, 1.0, 2.0, 3.0, 4.0], dtype=float)

# y must have same length as x along axis 0
y = np.array([1.0, 2.0, 1.5, 3.0, 2.5], dtype=float)

# w (weights) is optional, must be positive
w = np.array([1.0, 1.0, 1.0, 1.0, 1.0], dtype=float)

ARGS = {
    'x': x,
    'y': y,
    'w': w,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
hampel_filter Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/filtering.py
code served (permissive licence) · get_code("0dcf9273d4f1954c")
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 1, dtype family float, shape (15,) 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 sample 1D numpy array (column of data) with some outliers
np.random.seed(0)
col = np.array([1.0, 1.1, 1.2, 10.0, 1.3, 1.2, 1.1, 1.0, 1.1, 1.2, 50.0, 1.3, 1.2, 1.1, 1.0], dtype=np.float64)

ARGS = {
    "col": col,
    "window_size": 7,
    "n_sigma": 2
}
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 1, dtype family float, shape (15,) — shape (15,) 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
indices_of_first_last_non_nan_chunks Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/triangulation.py
code served (permissive licence) · get_code("45ac922f73ed0050")
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, ndim 1, dtype family int; 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 numpy as np
import pandas as pd

# Create a pandas Series with some NaN values and chunks of non-NaN values
np.random.seed(0)
data = np.array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0,
                 np.nan, np.nan, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0, 17.0, 18.0,
                 19.0, 20.0, np.nan, np.nan, np.nan])
series = pd.Series(data)

ARGS = {
    'series': series,
    'min_chunk_size': 10,
    'chunk_choice_method': 'largest',
    'trim_output_chunk': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family int — a tuple 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
is_image_file Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/common.py
code served (permissive licence) · get_code("728c82d2ed4d8df5")
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.

from pathlib import Path
import tempfile
import os

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

# Create a real image file with .png extension
png_file = os.path.join(temp_dir, "test_image.png")
with open(png_file, 'w') as f:
    f.write("fake image data")

# Create a non-image file
non_image_file = os.path.join(temp_dir, "test_file.txt")
with open(non_image_file, 'w') as f:
    f.write("not an image")

# Test with an actual image file path
ARGS = {
    "path": png_file
}
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 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
is_right Ran
ran, violated the contract we drafted
zju3dv/mvpose/src/m_utils/evaluate.py
code served (permissive licence) · get_code("ac6857abab524ac7")
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 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 numpy as np

# Create small numpy arrays representing 3D points (bone endpoints)
# These represent coordinates in 3D space
model_start_point = np.array([0.0, 0.0, 0.0])
model_end_point = np.array([1.0, 1.0, 1.0])
gt_strat_point = np.array([0.1, 0.1, 0.1])
gt_end_point = np.array([1.1, 1.1, 1.1])
alpha = 0.5

ARGS = {
    "model_start_point": model_start_point,
    "model_end_point": model_end_point,
    "gt_strat_point": gt_strat_point,
    "gt_end_point": gt_end_point,
    "alpha": alpha,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — 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
is_video_file Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/common.py
code served (permissive licence) · get_code("7b1c01d6703e3ce6")
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 tempfile
import os
from pathlib import Path

# Create a temporary directory with a test video file
temp_dir = tempfile.mkdtemp()
video_file_path = os.path.join(temp_dir, "test_video.mp4")

# Create an empty file with .mp4 extension
Path(video_file_path).touch()

ARGS = {
    "path": video_file_path
}
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 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
numpify Ran
ran; our contract draft was wrong, not the code
zju3dv/mvpose/src/m_utils/evaluate.py
code served (permissive licence) · get_code("0637e1b8059b1eed")
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 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 torch
import numpy as np

torch.manual_seed(0)

# Construct the info_dicts structure that numpify expects
# The function iterates over info_dicts.values(), and each info_dict has:
# - 'image_data': a tensor that needs squeeze() and numpy()
# - A key at index [0] that contains a list of persons
# - Each person has 'heatmap_data' and 'cropped_img' tensors

person_1 = {
    'heatmap_data': torch.randn(1, 4, 5),  # Will squeeze to (4, 5)
    'cropped_img': torch.randn(1, 3, 6)    # Will squeeze to (3, 6)
}

person_2 = {
    'heatmap_data': torch.randn(1, 4, 5),
    'cropped_img': torch.randn(1, 3, 6)
}

info_dict_1 = {
    'image_data': torch.randn(1, 7, 8),  # Will squeeze to (7, 8)
    0: [person_1, person_2]
}

person_3 = {
    'heatmap_data': torch.randn(1, 4, 5),
    'cropped_img': torch.randn(1, 3, 6)
}

info_dict_2 = {
    'image_data': torch.randn(1, 7, 8),
    0: [person_3]
}

info_dicts = {
    'sample_1': info_dict_1,
    'sample_2': info_dict_2
}

ARGS = {
    'info_dicts': info_dicts
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — 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
one_euro_filter_1d Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/filtering.py
code served (permissive licence) · get_code("d785e8aeb6422c48")
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 Series. 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
import pandas as pd

# Create a simple pandas Series with some values and NaNs
np.random.seed(0)
data = np.array([1.0, 1.5, 2.0, np.nan, 3.0, 3.2, 3.5, 4.0, np.nan, 4.5, 5.0])
col = pd.Series(data)

# Create the config_dict with the nested structure expected by the function
config_dict = {
    'filtering': {
        'one_euro': {
            'cut_off_frequency': 2.5,
            'beta': 0.9,
            'd_cut_off_frequency': 1.0
        }
    }
}

# Frame rate in Hz
frame_rate = 30

ARGS = {
    'config_dict': config_dict,
    'frame_rate': frame_rate,
    'col': col
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Series — a Series output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Series output the recorded ndim 1 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
persons_combinations Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/personAssociation.py
code served (permissive licence) · get_code("8b1697fa0742ed3c")
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, 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 json
import tempfile
import os

# Create temporary JSON files with person detection data
temp_dir = tempfile.mkdtemp()

# Camera 0: 2 persons detected
json_file_0 = os.path.join(temp_dir, "cam0.json")
with open(json_file_0, 'w') as f:
    json.dump({
        "people": [
            {"pose_keypoints_2d": [1.0, 2.0, 0.9, 3.0, 4.0, 0.8, 5.0, 6.0, 0.7]},
            {"pose_keypoints_2d": [10.0, 11.0, 0.95, 12.0, 13.0, 0.85, 14.0, 15.0, 0.75]}
        ]
    }, f)

# Camera 1: 1 person detected
json_file_1 = os.path.join(temp_dir, "cam1.json")
with open(json_file_1, 'w') as f:
    json.dump({
        "people": [
            {"pose_keypoints_2d": [20.0, 21.0, 0.92, 22.0, 23.0, 0.82, 24.0, 25.0, 0.72]}
        ]
    }, f)

# Camera 2: no persons detected (all NaN keypoints)
json_file_2 = os.path.join(temp_dir, "cam2.json")
with open(json_file_2, 'w') as f:
    json.dump({
        "people": [
            {"pose_keypoints_2d": [np.nan, np.nan, 0.0, np.nan, np.nan, 0.0, np.nan, np.nan, 0.0]}
        ]
    }, f)

ARGS = {
    "json_files_framef": [json_file_0, json_file_1, json_file_2]
}
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, 3) — shape (2, 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
recursive_update Ran
ran on a synthesized input
perfanalytics/pose2sim/Pose2Sim/Pose2Sim.py
code served (permissive licence) · get_code("b05140de9b16a9d2")
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 torch
import numpy as np

torch.manual_seed(0)

# Set up arguments for recursive_update function
dict_to_update = {
    'model': {
        'learning_rate': 0.001,
        'epochs': 10,
        'layers': {
            'hidden_dim': 64,
            'activation': 'relu'
        }
    },
    'data': {
        'batch_size': 32,
        'shuffle': True
    }
}

dict_with_new_values = {
    'model': {
        'learning_rate': 0.002,
        'layers': {
            'activation': 'gelu'
        }
    },
    'data': {
        'shuffle': False
    },
    'new_key': 'new_value'
}

ARGS = {
    'dict_to_update': dict_to_update,
    'dict_with_new_values': dict_with_new_values
}
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 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
create_textbox Not yet run
unverified (harvested, no recorded run)
perfanalytics/pose2sim/Pose2Sim/synchronization.py
code served (permissive licence) · get_code("62d5da938b7c82fb")
How it ran

no invocation recorded.

keypoints_ui Not yet run
unverified (harvested, no recorded run)
perfanalytics/pose2sim/Pose2Sim/synchronization.py
code served (permissive licence) · get_code("86773419dd277668")
How it ran

no invocation recorded.

retrieve_right_trc_order Not yet run
unverified (harvested, no recorded run)
perfanalytics/pose2sim/Pose2Sim/triangulation.py
code served (permissive licence) · get_code("374a6ef6d235197a")
How it ran

no invocation recorded.

setup_backend_device Not yet run
unverified (harvested, no recorded run)
perfanalytics/pose2sim/Pose2Sim/poseEstimation.py
code served (permissive licence) · get_code("754bc9064a046c1d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 19 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 19 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

This paper addresses the problem of 3D pose estimation for multiple people in a few calibrated camera views. The main challenge of this problem is to find the cross-view correspondences among noisy and incomplete 2D pose predictions. Most previous methods address this challenge by directly reasoning in 3D using a pictorial structure model, which is inefficient due to the huge state space. We propose a fast and robust approach to solve this problem. Our key idea is to use a multi-way matching algorithm to cluster the detected 2D poses in all views. Each resulting cluster encodes 2D poses of the same person across different views and consistent correspondences across the keypoints, from which the 3D pose of each person can be effectively inferred. The proposed convex optimization based multi-way matching algorithm is efficient and robust against missing and false detections, without knowing the number of people in the scene. Moreover, we propose to combine geometric and appearance cues for cross-view matching. The proposed approach achieves significant performance gains from the state-of-the-art (96.3% vs. 90.6% and 96.9% vs. 88% on the Campus and Shelf datasets, respectively), while being efficient for real-time applications.

For agents

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

get_harvested_code_for_paper("1901.04111")
get_code_for_paper("1901.04111")
have("1901.04111")

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

curl https://syntology.ai/api/ran/1901.04111.json

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

[![Syntology run record](https://syntology.ai/api/ran/1901.04111.svg)](https://syntology.ai/paper/1901.04111)

Connect an agent — have() is free.