SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2302.01058 · AAAI · 2023

IKOL: Inverse Kinematics Optimization Layer for 3D Human Pose and Shape Estimation via Gauss-Newton Differentiation

Yuexin Ma, Jingyi Yu, Ye Shi, Jingya Wang, Lan Xu, Juze Zhang

arXiv · PDF · Open in the Atlas

Code that ran

7 of 7 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, 0 violated it and 6 have no contract verdict; 1 failure of our instrument; 0 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/2302.01058.json · badge: /api/ran/2302.01058.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
Juzezhang/IKOL canonical (classifier label)
classifier basis not recorded on the link
7 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_similarity_transform Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Juzezhang/IKOL/ikol/core/evaluation/mesh_eval.py
code served (permissive licence) · get_code("7a9043e33d67fc2d")
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 (5, 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

# Create small test point clouds
# N=5 points, 3D coordinates
np.random.seed(0)
source_points = np.random.randn(5, 3).astype(np.float32)
target_points = np.random.randn(5, 3).astype(np.float32)

ARGS = {
    "source_points": source_points,
    "target_points": target_points,
}
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 (5, 3) — shape (5, 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
keypoint_accel_error Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/core/evaluation/eval_utils.py
code served (permissive licence) · get_code("a34a91ed922317b3")
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 (0,) 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 ground truth keypoints: shape (N, 14, 3)
# N=5 frames, 14 keypoints, 3 coordinates (x, y, z)
gt = np.random.randn(5, 14, 3).astype(np.float32)

# Create predicted keypoints: same shape as gt
pred = np.random.randn(5, 14, 3).astype(np.float32)

# Create mask: shape (N,) - boolean array indicating visibility
# True = visible, False = invisible
mask = np.array([True, True, False, True, True], dtype=bool)

ARGS = {
    "gt": gt,
    "pred": pred,
    "mask": mask,
}
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 (0,) — shape (0,) 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
keypoint_mpjpe Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/core/evaluation/eval_utils.py
code served (permissive licence) · get_code("70b980eea5c0c4af")
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 float32, 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

def compute_similarity_transform(pred, gt):
    """Helper function for procrustes alignment."""
    # Simple stand-in: just return pred unchanged for testing
    return pred

# Set random seed for reproducibility
np.random.seed(0)

# Create small test data
batch_size = 2
num_keypoints = 4
keypoint_dims = 3

# Predicted keypoints: shape (N, K, C) = (2, 4, 3)
pred = np.random.randn(batch_size, num_keypoints, keypoint_dims).astype(np.float32)

# Ground truth keypoints: shape (N, K, C) = (2, 4, 3)
gt = np.random.randn(batch_size, num_keypoints, keypoint_dims).astype(np.float32)

# Mask for visibility: shape (N, K) = (2, 4)
# All True to ensure mask.any() passes
mask = np.ones((batch_size, num_keypoints), dtype=bool)

# Alignment method
alignment = 'none'

ARGS = {
    'pred': pred,
    'gt': gt,
    'mask': mask,
    'alignment': alignment
}
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 float32, 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
norm_heatmap Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/models/IKOL24.py
code served (permissive licence) · get_code("a828d7ddc0f76aac")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 4, dtype family float, shape (2, 3, 4, 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)

# Create a small heatmap tensor with shape (batch, channels, height, width)
# Using distinct dimensions: batch=2, channels=3, height=4, width=5
heatmap = torch.randn(2, 3, 4, 5)

ARGS = {
    'norm_type': 'softmax',
    'heatmap': heatmap
}
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, 3, 4, 5) — shape (2, 3, 4, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
vertice_pve Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/core/evaluation/eval_utils.py
code served (permissive licence) · get_code("ff9bfb70377fb976")
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 float32, 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
import torch

torch.manual_seed(0)

# Create small sample vertex data
# Batch size: 2, number of vertices: 4, coordinates: 3 (x, y, z)
batch_size = 2
num_verts = 4
coord_dim = 3

# Create predicted vertices: (2, 4, 3)
pred_verts = np.random.randn(batch_size, num_verts, coord_dim).astype(np.float32)

# Create target vertices: (2, 4, 3)
target_verts = np.random.randn(batch_size, num_verts, coord_dim).astype(np.float32)

ARGS = {
    "pred_verts": pred_verts,
    "target_verts": target_verts,
}
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 float32, 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
vis_keypoints Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/models/IKOL24.py
code served (permissive licence) · get_code("ec26782c7a888141")
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 3, dtype family int, shape (100, 100, 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 cv2

# Create a small test image (100x100 RGB)
img = np.random.randint(0, 255, (100, 100, 3), dtype=np.uint8)

# Create keypoints array: shape (3, num_keypoints)
# Format: [x_coords, y_coords, confidence_scores]
num_keypoints = 5
kps = np.array([
    [10, 20, 30, 40, 50],      # x coordinates
    [15, 25, 35, 45, 55],      # y coordinates
    [0.9, 0.8, 0.5, 0.3, 0.7]  # confidence scores
], dtype=np.float32)

# Define keypoint connections as pairs of indices
kps_lines = [
    [0, 1],
    [1, 2],
    [2, 3],
    [3, 4]
]

ARGS = {
    'img': img,
    'kps': kps,
    'kps_lines': kps_lines,
    'kp_thresh': 0.4,
    'alpha': 0.5,
    'thickness': 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, matplotlib, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (100, 100, 3) — shape (100, 100, 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
weighted_l1_loss Ran
ran on a synthesized input
Juzezhang/IKOL/ikol/models/criterion.py
code served (permissive licence) · get_code("3b94245b39283e5c")
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 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

torch.manual_seed(0)

# Create small tensors with distinct shapes for input and target
input_tensor = torch.randn(2, 4)  # batch_size=2, dim=4
target_tensor = torch.randn(2, 4)  # same shape as input

# Create weights tensor - same shape as input/target for element-wise multiplication
weights_tensor = torch.ones(2, 4)

# Set size_average to True
size_average = True

ARGS = {
    "input": input_tensor,
    "target": target_tensor,
    "weights": weights_tensor,
    "size_average": size_average,
}
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 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

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 3 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 2 later) and 4 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
compute_similarity_transform 1 other paper
1 dated earlier
2108.12617 (2021-08)
AP-10K: A Benchmark for Animal Pose Estimation in the Wild
norm_heatmap 1 other paper
1 later
2304.05690 (2023-04)
HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-body Mesh Recovery
weighted_l1_loss 1 other paper
1 later
2304.05690 (2023-04)
HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-body Mesh Recovery

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 presents an inverse kinematic optimization layer (IKOL) for 3D human pose and shape estimation that leverages the strength of both optimization-and regression-based methods within an end-to-end framework. IKOL involves a nonconvex optimization that establishes an implicit mapping from an image's 3D keypoints and body shapes to the relative body-part rotations. The 3D keypoints and the body shapes are the inputs and the relative body-part rotations are the solutions. However, this procedure is implicit and hard to make differentiable. So, to overcome this issue, we designed a Gauss-Newton differentiation (GN-Diff) procedure to differentiate IKOL. GN-Diff iteratively linearizes the nonconvex objective function to obtain Gauss-Newton directions with closed form solutions. Then, an automatic differentiation procedure is directly applied to generate a Jacobian matrix for end-to-end training. Notably, the GN-Diff procedure works fast because it does not rely on a time-consuming implicit differentiation procedure. The twist rotation and shape parameters are learned from the neural networks and, as a result, IKOL has a much lower computational overhead than most existing optimization-based methods. Additionally, compared to existing regression-based methods, IKOL provides a more accurate mesh-image correspondence. This is because it iteratively reduces the distance between the keypoints and also enhances the reliability of the pose structures. Extensive experiments demonstrate the superiority of our proposed framework over a wide range of 3D human pose and shape estimation methods. Code is available at https://github.com/Juzezhang/IKOL.

For agents

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

get_harvested_code_for_paper("2302.01058")
get_code_for_paper("2302.01058")
have("2302.01058")

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

curl https://syntology.ai/api/ran/2302.01058.json

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

[![Syntology run record](https://syntology.ai/api/ran/2302.01058.svg)](https://syntology.ai/paper/2302.01058)

Connect an agent — have() is free.