SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.06819 · ICCV · 2021

EventHPE: Event-based 3D Human Pose and Shape Estimation

Sen Wang, Chuan Guo, Li Cheng, Pengyu Wang, Xinxin Zuo, Shihao Zou, Minglun Gong, Xiaoqin Hu, Shoushun Chen

arXiv · PDF · Open in the Atlas

Code that ran

10 of 13 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 7 failures of our instrument; 3 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/2108.06819.json · badge: /api/ran/2108.06819.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
jimmyzou/eventhpe marked official in Papers with Code
Papers with Code archive
10 ran · 3 no recorded run
FunctionStatusWhere it lives
Regressor Ran
ran on a synthesized input · constructed an object; no result was computed
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("57a649d3f633b7ee")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Regressor: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The Regressor.__init__ takes one parameter: pose_dim
# The forward method expects input x of shape (B, T, F) where F=2048
# So we need to construct a valid input tensor

# Create a small input tensor: (B=2, T=3, F=2048)
x = torch.randn(2, 3, 2048)

ARGS = {
    'pose_dim': 24 * 6
}
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 Regressor — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Regressor 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
TemporalEncoder Ran
ran on a synthesized input · constructed an object; no result was computed
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("b1ed1f7bf1a9fa18")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, TemporalEncoder: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# The forward method uses:
# - x: tensor with shape (B, T, F) where F=2048
# - hidden_feats: tensor that gets unsqueezed to (1, B, F_hidden)
# The GRU input_size is hardcoded to 2048
# hidden_feats should have shape (B, something) - needs to match h0_size after repeat

# Let's construct valid inputs
batch_size = 2
seq_length = 3
feature_size = 2048
hidden_size = 2048

# Create input tensor x: (B, T, F) = (2, 3, 2048)
x = torch.randn(batch_size, seq_length, feature_size)

# Create hidden_feats: (B, hidden_size) 
# After unsqueeze(0) it becomes (1, B, hidden_size)
# Then repeat(h0_size, 1, 1) needs h0_size to match
# h0_size = n_layers * (bidirectional + 1)
# For n_layers=1, bidirectional=False: h0_size = 1
hidden_feats = torch.randn(batch_size, hidden_size)

ARGS = {
    "n_layers": 1,
    "hidden_size": 2048,
    "add_linear": False,
    "bidirectional": False,
    "use_residual": 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 TemporalEncoder — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a TemporalEncoder 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
VIBERegressor Ran
ran on a synthesized input
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("af15be9cb121f659")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) 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 torch
import torch.nn as nn
import numpy as np
import torch.nn.functional as F
import tempfile
import os

# Create a temporary file with mock SMPL mean parameters
temp_dir = tempfile.mkdtemp()
smpl_mean_params_path = os.path.join(temp_dir, 'smpl_mean_params.npy')

pose_dim = 24 * 6  # 144
mean_params = {
    'pose': np.random.randn(pose_dim).astype(np.float32)
}
np.save(smpl_mean_params_path, mean_params)

INIT_ARGS = {
    'pose_dim': 144,
    'smpl_mean_params': smpl_mean_params_path
}

# Forward expects input of shape (B, T, F) where F = 2048 + pose_dim + 3 is NOT correct
# Looking at the forward: x has shape (B, T, F), then after view it's (B*T, F)
# Then it's concatenated with pred_pose (B*T, pose_dim) and pred_trans (B*T, 3)
# So xc has shape (B*T, F + pose_dim + 3)
# fc1 expects input of size 2048 + pose_dim + 3
# Therefore F must be 2048

B = 2
T = 3
F = 2048

ARGS = {
    'x': torch.randn(B, T, F),
    'n_iter': 3
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
numpy, torch — 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
batch_rodrigues Ran
ran; our contract draft was wrong, not the code
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("ea9d895836659ed3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# batch_rodrigues takes theta: a tensor of shape [B, 3]
# where B is batch size and 3 is the axis-angle representation dimension
theta = torch.randn(2, 3)

ARGS = {
    "theta": theta
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
cross_product Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("54d29d910459611f")
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, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function expects two 3D vectors in batched form
# u and v should be tensors of shape (batch_size, 3)
# where batch_size is the number of vector pairs

batch_size = 2
u = torch.randn(batch_size, 3)
v = torch.randn(batch_size, 3)

ARGS = {
    "u": u,
    "v": v,
}
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, 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
delta_rotmat_to_rotmat Ran
ran; our contract draft was wrong, not the code
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("b93e708054daf095")
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 5, dtype family float, shape (2, 3, 24, 3, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Analyze the function:
# - init_rotmats: reshaped to [-1, 3, 3], so shape should be (B*24, 3, 3) or similar
# - delta_rotmats: indexed as [:, i, :, :, :] where i goes 0 to num_steps-1
#   then reshaped to [-1, 3, 3], so shape is (B, num_steps, ?, 3, 3)
# - num_steps: integer, used in range()

B = 2
num_steps = 3
num_joints = 24

# init_rotmats should be (B, 24, 3, 3) based on the reshape and later reshape back
init_rotmats = torch.randn(B, num_joints, 3, 3)

# delta_rotmats should be (B, num_steps, num_joints, 3, 3)
# When indexed as [:, i, :, :, :] it gives (B, num_joints, 3, 3)
# Then reshaped to [-1, 3, 3] gives (B*num_joints, 3, 3)
delta_rotmats = torch.randn(B, num_steps, num_joints, 3, 3)

ARGS = {
    "init_rotmats": init_rotmats,
    "delta_rotmats": delta_rotmats,
    "num_steps": num_steps,
}
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 5, dtype family float, shape (2, 3, 24, 3, 3) — shape (2, 3, 24, 3, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalize_vector Ran
ran; our contract draft was wrong, not the code
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("7bea3b3b5e1cf218")
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 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 vector tensor with batch size 2 and dimension 4
v = torch.randn(2, 4, requires_grad=True)

ARGS = {
    "v": v,
    "return_mag": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
projection_torch Ran
ran; our contract draft was wrong, not the code
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("b6567d5d4633ec0f")
How it ran

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

torch.manual_seed(0)

# Create xyz tensor: shape (2, 4, 3) - batch_size=2, sequence_len=4, coords=3
xyz = torch.randn(2, 4, 3)

# Create intr_param with 4 intrinsic parameters: [fx, fy, cx, cy]
intr_param = torch.tensor([500.0, 500.0, 320.0, 240.0])

# H and W are image height and width
H = 480.0
W = 640.0

ARGS = {
    'xyz': xyz,
    'intr_param': intr_param,
    'H': H,
    'W': W
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 2) — shape (2, 4, 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
quat2mat Ran
ran; our contract draft was wrong, not the code
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("baa1c1737c3796df")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function quat2mat expects a quaternion tensor of shape [B, 4]
# where B is batch size and 4 represents (w, x, y, z) components
# Create a small batch of 2 quaternions
quat = torch.randn(2, 4)

ARGS = {
    "quat": quat
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rot6d_to_rotmat Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("fecaa96c9563582c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small batch of 6D orthogonal representations
# The function expects input shape (batch_size, 6)
batch_size = 2
ortho6d = torch.randn(batch_size, 6)

ARGS = {
    "ortho6d": ortho6d
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
EventTrackNet Not yet run
unverified (harvested, no recorded run)
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("1f9e5237fb7e7a54")
How it ran

no invocation recorded.

SMPL Not yet run
unverified (harvested, no recorded run)
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("4716d386941e0081")
How it ran

no invocation recorded.

batch_global_rigid_transformation Not yet run
unverified (harvested, no recorded run)
jimmyzou/eventhpe/event_pose_estimation/model.py
code served (permissive licence) · get_code("f4e4b5b0cdf4737d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 3 later) and 10 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
quat2mat
brick of this paper's assembly: helper
7 other papers
3 dated earlier, 4 later
1905.03244 (2019-05)
Convolutional Mesh Regression for Single-Image Human Shape Reconstruction
2004.01946 (2020-04)
Weakly-Supervised Mesh-Convolutional Hand Reconstruction in the Wild
2006.05734 (2020-06)
3D Human Mesh Regression with Dense Correspondence
and 4 more not named here
batch_rodrigues
brick of this paper's assembly: helper
3 other papers
3 later
aaai_28166 (2024 (year only))
Keypoint Fusion for RGB-D Based 3D Hand Pose Estimation
2402.17062 (2024-02)
HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields
2403.07912 (2024-03)
HandGCAT: Occlusion-Robust 3D Hand Mesh Reconstruction from Monocular Images
cross_product
brick of this paper's assembly: helper
3 other papers
1 dated earlier, 2 later · 1 only as an assembly brick
2106.12782 (2021-06)
Hamiltonian-based Neural ODE Networks on the SE(3) Manifold For Dynamics Learning and Control
2209.06407 (2022-09)
Viewer-Centred Surface Completion for Unsupervised Domain Adaptation in 3D Object Detection
2303.08240 (2023-03)
Parametric Surface Constrained Upsampler Network for Point Cloud

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

Event camera is an emerging imaging sensor for capturing dynamics of moving objects as events, which motivates our work in estimating 3D human pose and shape from the event signals. Events, on the other hand, have their unique challenges: rather than capturing static body postures, the event signals are best at capturing local motions. This leads us to propose a two-stage deep learning approach, called EventHPE. The first-stage, FlowNet, is trained by unsupervised learning to infer optical flow from events. Both events and optical flow are closely related to human body dynamics, which are fed as input to the ShapeNet in the second stage, to estimate 3D human shapes. To mitigate the discrepancy between image-based flow (optical flow) and shape-based flow (vertices movement of human body shape), a novel flow coherence loss is introduced by exploiting the fact that both flows are originated from the identical human motion. An in-house event-based 3D human dataset is curated that comes with 3D pose and shape annotations, which is by far the largest one to our knowledge. Empirical evaluations on DHP19 dataset and our in-house dataset demonstrate the effectiveness of our approach.

For agents

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

get_harvested_code_for_paper("2108.06819")
get_code_for_paper("2108.06819")
have("2108.06819")

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

curl https://syntology.ai/api/ran/2108.06819.json

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

[![Syntology run record](https://syntology.ai/api/ran/2108.06819.svg)](https://syntology.ai/paper/2108.06819)

Connect an agent — have() is free.