SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1902.09212 · 2019

Deep High-Resolution Representation Learning for Human Pose Estimation

arXiv · PDF · Open in the Atlas

Code that ran

16 of 25 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; 1 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 8 failures of our instrument; 9 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/1902.09212.json · badge: /api/ran/1902.09212.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
Microsoft/human-pose-estimation.pytorch canonical (classifier label)
classifier basis: plan_row
8 ran · 1 no recorded run
mks0601/PoseFix_RELEASE canonical (classifier label)
classifier basis: census
0 ran · 4 no recorded run
wsjzha/deep-high-resolution-net.pytorch canonical (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
Vill-Lab/2022-TIP-HCGA community (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
HRNet/HRNet-MaskRCNN-Benchmark extension (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
visionNoob/hrnet_pytorch reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
calc_dists Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Microsoft/human-pose-estimation.pytorch/lib/core/evaluate.py
code served (permissive licence) · get_code("b91b38261b03587d")
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 (3, 2) 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

# Construct valid arguments for calc_dists function
# The function expects:
# - preds: array of shape (batch, channels, 2) with predicted coordinates
# - target: array of shape (batch, channels, 2) with target coordinates
# - normalize: array of shape (batch,) with normalization factors

batch_size = 2
num_channels = 3

# Create preds: (batch=2, channels=3, coords=2)
preds = np.random.rand(batch_size, num_channels, 2).astype(np.float32) * 100

# Create target: (batch=2, channels=3, coords=2)
# Use values > 1 to satisfy the condition in the function
target = np.random.rand(batch_size, num_channels, 2).astype(np.float32) * 100 + 2

# Create normalize: (batch=2,) with positive values for normalization
normalize = np.array([10.0, 15.0], dtype=np.float32)

ARGS = {
    'preds': preds,
    'target': target,
    'normalize': normalize
}
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 (3, 2) — shape (3, 2) 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
calc_iou Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
HRNet/HRNet-MaskRCNN-Benchmark/tools/test_net.py
code served (permissive licence) · get_code("451bcfd505206dce")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned float, 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 numpy as np

torch.manual_seed(0)

# The function calc_iou expects two bounding boxes in format [x1, y1, x2, y2]
# where (x1, y1) is top-left corner and (x2, y2) is bottom-right corner

# Create two overlapping bounding boxes
a = [10, 20, 50, 60]  # Box a: x1=10, y1=20, x2=50, y2=60
b = [30, 40, 70, 80]  # Box b: x1=30, y1=40, x2=70, y2=80

ARGS = {
    'a': a,
    'b': b
}
Setup block drafted by
a language model, Claude Sonnet 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 float, 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
conv3x3 Ran
ran; our contract draft was wrong, not the code
Microsoft/human-pose-estimation.pytorch/lib/models/pose_resnet.py
pointer only (licence: MIT); the stored copy get_code serves, from lixiaotong97/DSU, is pointer only (licence: NONE) · get_code("fac5364e2f53c6db")
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 lixiaotong97/DSU, 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. A run is recorded.

cpu_nms_wrapper Ran
ran on a synthesized input
Microsoft/human-pose-estimation.pytorch/lib/nms/nms.py
code served (permissive licence) · get_code("0bdca114d99f4177")
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 function. 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

def cpu_nms(dets, thresh):
    """
    Dummy implementation of CPU NMS for testing purposes.
    dets: array of shape (N, 5) with columns [x1, y1, x2, y2, score]
    thresh: IoU threshold
    Returns: array of indices to keep
    """
    if len(dets) == 0:
        return np.array([], dtype=np.int32)
    
    x1 = dets[:, 0]
    y1 = dets[:, 1]
    x2 = dets[:, 2]
    y2 = dets[:, 3]
    scores = dets[:, 4]
    
    areas = (x2 - x1 + 1) * (y2 - y1 + 1)
    order = scores.argsort()[::-1]
    
    keep = []
    while order.size > 0:
        i = order[0]
        keep.append(i)
        
        if order.size == 1:
            break
        
        xx1 = np.maximum(x1[i], x1[order[1:]])
        yy1 = np.maximum(y1[i], y1[order[1:]])
        xx2 = np.minimum(x2[i], x2[order[1:]])
        yy2 = np.minimum(y2[i], y2[order[1:]])
        
        w = np.maximum(0.0, xx2 - xx1 + 1)
        h = np.maximum(0.0, yy2 - yy1 + 1)
        inter = w * h
        
        ovr = inter / (areas[i] + areas[order[1:]] - inter)
        inds = np.where(ovr <= thresh)[0]
        order = order[inds + 1]
    
    return np.array(keep, dtype=np.int32)


torch.manual_seed(0)
np.random.seed(0)

# Create test detections: (N, 5) where columns are [x1, y1, x2, y2, score]
dets = np.array([
    [10.0, 20.0, 100.0, 200.0, 0.9],
    [15.0, 25.0, 105.0, 205.0, 0.8],
], dtype=np.float32)

thresh = 0.5

ARGS = {
    "thresh": thresh,
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
dist_acc Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Microsoft/human-pose-estimation.pytorch/lib/core/evaluate.py
code served (permissive licence) · get_code("1c2da88027e6e151")
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 float64, 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 a sample distance array with some valid values and some -1 values
# The function expects an array where -1 indicates "ignore this value"
dists = np.array([0.2, 0.6, -1, 0.3, 0.8, -1, 0.1, 0.4])

# Create arguments dictionary
ARGS = {
    "dists": dists,
    "thr": 0.5
}
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 float64, 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
euclidean_dist Ran
ran; our contract draft was wrong, not the code
Vill-Lab/2022-TIP-HCGA/layers/triplet_loss.py
pointer only (licence: MIT); the stored copy get_code serves, from needylove/regression_tightness, is pointer only (licence: NONE) · get_code("c83d9f7fca950b77")
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 needylove/regression_tightness, 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. A run is recorded.

get_max_preds Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Microsoft/human-pose-estimation.pytorch/lib/core/inference.py
code served (permissive licence) · get_code("55c58ef5f795e3d1")
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 batch of heatmaps
# Shape: (batch_size, num_joints, height, width)
batch_size = 2
num_joints = 3
height = 4
width = 5

# Create realistic heatmap data with values between 0 and 1
np.random.seed(0)
batch_heatmaps = np.random.rand(batch_size, num_joints, height, width).astype(np.float32)

ARGS = {
    'batch_heatmaps': batch_heatmaps
}
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_model_name Ran
ran on a synthesized input
Microsoft/human-pose-estimation.pytorch/lib/core/config.py
code served (permissive licence) · get_code("18f7aa997852bcc0")
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 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.

class Config:
    def __init__(self):
        self.MODEL = ModelConfig()

class ModelConfig:
    def __init__(self):
        self.NAME = 'pose_resnet'
        self.IMAGE_SIZE = [256, 384]
        self.EXTRA = ExtraConfig()

class ExtraConfig:
    def __init__(self):
        self.NUM_LAYERS = 50
        self.NUM_DECONV_FILTERS = [256, 256, 256]

cfg = Config()

ARGS = {
    'cfg': cfg
}
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 tuple, ndim 1, dtype family float — 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
gpu_nms_wrapper Ran
ran on a synthesized input
Microsoft/human-pose-estimation.pytorch/lib/nms/nms.py
code served (permissive licence) · get_code("167b862a31de0adf")
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 function. 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)

# gpu_nms_wrapper takes thresh (float) and device_id (int)
# The returned function _nms takes dets (detections)
# dets should be a tensor of shape (N, 5) typically: [x1, y1, x2, y2, score]
# for NMS operations

ARGS = {
    'thresh': 0.5,
    'device_id': 0
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
hard_example_mining Ran
ran on a synthesized input
Vill-Lab/2022-TIP-HCGA/layers/triplet_loss.py
code served (permissive licence) · get_code("93f2c0b80ea7f36e")
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 2, 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 torch

torch.manual_seed(0)

# Create a distance matrix of shape [N, N]
# N = 4 samples
N = 4
dist_mat = torch.randn(N, N)
dist_mat = torch.abs(dist_mat)  # Make distances positive

# Create labels - we need samples grouped by class
# For hard example mining to work properly, we need multiple samples per class
# Let's have 2 classes with 2 samples each
labels = torch.LongTensor([0, 0, 1, 1])

ARGS = {
    "dist_mat": dist_mat,
    "labels": labels,
    "return_inds": 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 tuple, ndim 2, dtype family float — 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
normalize Ran
ran; our contract draft was wrong, not the code
Vill-Lab/2022-TIP-HCGA/layers/triplet_loss.py
code served (permissive licence) · get_code("e3b2a83f52101f2d")
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 tensor with batch size 2 and feature dimension 4
x = torch.randn(2, 4)

ARGS = {
    "x": x,
    "axis": -1
}
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
pil_loader Ran
ran, honoured the contract we drafted
Vill-Lab/2022-TIP-HCGA/engine/clustering.py
code served (permissive licence) · get_code("1508ae9fb90b8de1")
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 Image. 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
from PIL import Image
import numpy as np

# Create a temporary directory and a test image file
temp_dir = tempfile.mkdtemp()
image_path = os.path.join(temp_dir, "test_image.jpg")

# Create a simple test image and save it
test_image_array = np.random.randint(0, 256, (100, 100, 3), dtype=np.uint8)
test_image = Image.fromarray(test_image_array, mode='RGB')
test_image.save(image_path)

ARGS = {
    "path": image_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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
preprocess_features Ran
ran on a synthesized input
Vill-Lab/2022-TIP-HCGA/engine/clustering.py
code served (permissive licence) · get_code("53f8b1e59f08664a")
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 sample feature data: N=2 samples, ndim=4 features
np.random.seed(0)
npdata = np.random.randn(2, 4).astype('float32')

ARGS = {
    'npdata': npdata,
    'norm': False,
    'pca': 256
}
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
py_nms_wrapper Ran
ran on a synthesized input
Microsoft/human-pose-estimation.pytorch/lib/nms/nms.py
code served (permissive licence) · get_code("8551f210abffd79e")
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 function. 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 sample detection boxes with format [x1, y1, x2, y2, score]
# We need at least a few boxes to test the NMS algorithm
dets = np.array([
    [10.0, 20.0, 100.0, 150.0, 0.9],   # high confidence box
    [15.0, 25.0, 105.0, 155.0, 0.8],   # overlapping box with lower confidence
    [200.0, 200.0, 300.0, 350.0, 0.85], # separate box
    [205.0, 205.0, 305.0, 355.0, 0.7],  # overlapping with the third box
], dtype=np.float32)

thresh = 0.3

ARGS = {
    "thresh": thresh
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
split_n Ran
ran on a synthesized input
Vill-Lab/2022-TIP-HCGA/deep_feature_clustersing.py
code served (permissive licence) · get_code("1be7fb0ef331f2a5")
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. 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)

# split_n takes a sequence and splits it into num roughly equal parts
# seq: any sequence-like object that supports len() and slicing
# num: number of splits to create

# Create a simple list as the sequence
seq = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

# Number of splits
num = 3

ARGS = {
    "seq": seq,
    "num": num
}
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 list — a list 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
voting Ran
ran; our contract draft was wrong, not the code
HRNet/HRNet-MaskRCNN-Benchmark/tools/test_net.py
code served (permissive licence) · get_code("bb37605b18846a6b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 torch
import sys
from multiprocessing import Manager

torch.manual_seed(0)

class BoundingBoxList:
    def __init__(self, bbox, size, extra_fields=None):
        self.bbox = bbox
        self.size = size  # (width, height)
        self.extra_fields = extra_fields if extra_fields is not None else {}
    
    def resize(self, size):
        # Return a new BoundingBoxList with the same data but different size
        new_bbox_list = BoundingBoxList(self.bbox.clone(), size, {k: v.clone() if isinstance(v, torch.Tensor) else v for k, v in self.extra_fields.items()})
        return new_bbox_list
    
    def transpose(self, method):
        # Flip horizontally (method=0)
        # For simplicity, just return a new instance with flipped bboxes
        width, height = self.size
        flipped_bbox = self.bbox.clone()
        # Flip x coordinates: new_x1 = width - old_x2, new_x2 = width - old_x1
        flipped_bbox[:, 0] = width - self.bbox[:, 2]
        flipped_bbox[:, 2] = width - self.bbox[:, 0]
        new_bbox_list = BoundingBoxList(flipped_bbox, self.size, {k: v.clone() if isinstance(v, torch.Tensor) else v for k, v in self.extra_fields.items()})
        return new_bbox_list

# Create predictions for multiple scales
# predictions is a list of lists, where each inner list contains BoundingBoxList objects for each image
num_scales = 4
num_images = 3

predictions = []
for scale_idx in range(num_scales):
    scale_predictions = []
    for img_idx in range(num_images):
        # Create 2-3 bounding boxes per image per scale
        num_boxes = 2 + (scale_idx + img_idx) % 2
        bboxes = torch.tensor([
            [10.0 + scale_idx * 5, 20.0 + scale_idx * 5, 50.0 + scale_idx * 5, 60.0 + scale_idx * 5],
            [100.0 + img_idx * 10, 120.0 + img_idx * 10, 150.0 + img_idx * 10, 170.0 + img_idx * 10],
        ][:num_boxes])
        
        labels = torch.tensor([1.0, 2.0][:num_boxes])
        scores = torch.tensor([0.9 - scale_idx * 0.1, 0.8 - scale_idx * 0.1][:num_boxes])
        
        # Size varies by scale
        size = (640 + scale_idx * 100, 480 + scale_idx * 80)
        
        bbox_list = BoundingBoxList(
            bboxes,
            size,
            {'labels': labels, 'scores': scores}
        )
        scale_predictions.append(bbox_list)
    predictions.append(scale_predictions)

ARGS = {
    'predictions': predictions,
    'local_rank': 0,
    'flip': False,
    'num_process': 2,
    'overlap': 0.7
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch, tqdm — 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
aggregate_batch Not yet run
unverified (harvested, no recorded run)
mks0601/PoseFix_RELEASE/lib/tfflat/net_utils.py
code served (permissive licence) · get_code("d9b08c7049915445")
How it ran

no invocation recorded.

average_gradients Not yet run
unverified (harvested, no recorded run)
mks0601/PoseFix_RELEASE/lib/tfflat/net_utils.py
code served (permissive licence) · get_code("94ffb5d47eee6270")
How it ran

no invocation recorded.

clustering Not yet run
unverified (harvested, no recorded run)
Vill-Lab/2022-TIP-HCGA/deep_feature_clustersing.py
code served (permissive licence) · get_code("5d13a519b4ca488b")
How it ran

no invocation recorded.

get_pose_net Not yet run
unverified (harvested, no recorded run)
wsjzha/deep-high-resolution-net.pytorch/lib/models/pose_hrnet.py
code served (permissive licence) · get_code("2534c81208f21998")
How it ran

no invocation recorded.

get_pose_net Not yet run
unverified (harvested, no recorded run)
visionNoob/hrnet_pytorch/lib/models/pose_hrnet.py
code served (permissive licence) · get_code("78015ae361e585ca")
How it ran

no invocation recorded.

get_pose_net Not yet run
unverified (harvested, no recorded run)
Microsoft/human-pose-estimation.pytorch/lib/models/pose_resnet.py
code served (permissive licence) · get_code("9dfb17266423d20d")
How it ran

no invocation recorded.

get_variables_in_checkpoint_file Not yet run
unverified (harvested, no recorded run)
mks0601/PoseFix_RELEASE/lib/tfflat/saver.py
code served (permissive licence) · get_code("27cc68d97f7be674")
How it ran

no invocation recorded.

make_graph Not yet run
unverified (harvested, no recorded run)
Vill-Lab/2022-TIP-HCGA/engine/clustering.py
code served (permissive licence) · get_code("c1ac4ee52106027c")
How it ran

no invocation recorded.

sum_gradients Not yet run
unverified (harvested, no recorded run)
mks0601/PoseFix_RELEASE/lib/tfflat/net_utils.py
code served (permissive licence) · get_code("3eb931c10df5890d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 25 samples lifted from this paper's linked repositories, 25 carry a function identity and 0 do not. Of the 25 distinct functions with an identity, 17 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 1 the same month, 17 later) and 8 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
conv3x3 686 other papers
87 dated earlier, 5 the same month, 594 later · 31 only as an assembly brick
1406.2661 (2014-06)
Generative Adversarial Networks
1409.0575 (2014-09)
ImageNet Large Scale Visual Recognition Challenge
1412.6572 (2014-12)
Explaining and Harnessing Adversarial Examples
and 683 more not named here
pil_loader 66 other papers
3 dated earlier, 63 later
1502.03044 (2015-02)
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1711.02512 (2017-11)
Fine-tuning CNN Image Retrieval with No Human Annotation
and 63 more not named here
normalize 29 other papers
3 dated earlier, 26 later
1609.01775 (2016-09)
Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking
1804.01438 (2018-04)
Learning Discriminative Features with Multiple Granularities for Person Re-Identification
1811.07130 (2018-11)
Batch DropBlock Network for Person Re-identification and Beyond
and 26 more not named here
euclidean_dist 20 other papers
20 later · 4 only as an assembly brick
1903.07071 (2019-03)
Bag of Tricks and A Strong Baseline for Deep Person Re-identification
1904.02998 (2019-04)
Relation-Aware Global Attention for Person Re-identification
1904.07223 (2019-04)
Joint Discriminative and Generative Learning for Person Re-identification
and 17 more not named here
hard_example_mining 7 other papers
7 later
1903.07071 (2019-03)
Bag of Tricks and A Strong Baseline for Deep Person Re-identification
1904.07223 (2019-04)
Joint Discriminative and Generative Learning for Person Re-identification
2001.04193 (2020-01)
Deep Learning for Person Re-identification: A Survey and Outlook
and 4 more not named here
get_max_preds 6 other papers
6 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
1910.06278 (2019-10)
Distribution-Aware Coordinate Representation for Human Pose Estimation
and 3 more not named here
calc_dists 5 other papers
5 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
1910.06278 (2019-10)
Distribution-Aware Coordinate Representation for Human Pose Estimation
and 2 more not named here
dist_acc 5 other papers
5 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
1910.06278 (2019-10)
Distribution-Aware Coordinate Representation for Human Pose Estimation
and 2 more not named here
get_pose_net 5 other papers
5 later
1904.04514 (2019-04)
High-Resolution Representations for Labeling Pixels and Regions
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
and 2 more not named here
get_pose_net 5 other papers
5 later
1904.04514 (2019-04)
High-Resolution Representations for Labeling Pixels and Regions
1908.07919 (2019-08)
Deep High-Resolution Representation Learning for Visual Recognition
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
and 2 more not named here
cpu_nms_wrapper 2 other papers
2 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
gpu_nms_wrapper 2 other papers
2 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
py_nms_wrapper 2 other papers
2 later
1908.10357 (2019-08)
HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation
1908.11542 (2019-08)
Satellite Pose Estimation with Deep Landmark Regression and Nonlinear Pose Refinement
calc_iou 1 other paper
1 later
1904.04514 (2019-04)
High-Resolution Representations for Labeling Pixels and Regions
make_graph 1 other paper
1 later
2007.13467 (2020-07)
Identity-Guided Human Semantic Parsing for Person Re-Identification
preprocess_features 1 other paper
1 later
2007.13467 (2020-07)
Identity-Guided Human Semantic Parsing for Person Re-Identification
voting 1 other paper
1 later
1904.04514 (2019-04)
High-Resolution Representations for Labeling Pixels and Regions

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

More are linked than shown here.

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

Abstract

This is an official pytorch implementation of Deep High-Resolution Representation Learning for Human Pose Estimation. In this work, we are interested in the human pose estimation problem with a focus on learning reliable high-resolution representations. Most existing methods recover high-resolution representations from low-resolution representations produced by a high-to-low resolution network. Instead, our proposed network maintains high-resolution representations through the whole process. We start from a high-resolution subnetwork as the first stage, gradually add high-to-low resolution subnetworks one by one to form more stages, and connect the mutli-resolution subnetworks in parallel. We conduct repeated multi-scale fusions such that each of the high-to-low resolution representations receives information from other parallel representations over and over, leading to rich high-resolution representations. As a result, the predicted keypoint heatmap is potentially more accurate and spatially more precise. We empirically demonstrate the effectiveness of our network through the superior pose estimation results over two benchmark datasets: the COCO keypoint detection dataset and the MPII Human Pose dataset. The code and models have been publicly available at \url{https://github.com/leoxiaobin/deep-high-resolution-net.pytorch}.

For agents

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

get_harvested_code_for_paper("1902.09212")
get_code_for_paper("1902.09212")
have("1902.09212")

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

curl https://syntology.ai/api/ran/1902.09212.json

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

[![Syntology run record](https://syntology.ai/api/ran/1902.09212.svg)](https://syntology.ai/paper/1902.09212)

Connect an agent — have() is free.