SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.09805 · ECCV · 2024

On the Utility of 3D Hand Poses for Action Recognition

Angela Yao, Dibyadip Chatterjee, Fadime Sener, Shugao Ma, Salman Shamil

arXiv · PDF · Open in the Atlas

Code that ran

8 of 13 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (5 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 5 have no contract verdict; 2 failures of our instrument; 5 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/2403.09805.json · badge: /api/ran/2403.09805.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
s-shamil/HandFormer marked official in Papers with Code
Papers with Code archive
8 ran · 5 no recorded run
FunctionStatusWhere it lives
PositionalEncoding Ran
ran on a synthesized input · constructed an object; no result was computed
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("ea63108588c955b8")
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, PositionalEncoding: 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
import math
from torch.autograd import Variable

torch.manual_seed(0)

# PositionalEncoding.__init__ takes:
# - self
# - d_model: int (dimension of the model)
# - max_len: int (default 5000, max sequence length)

# PositionalEncoding.forward takes:
# - self
# - x: tensor with shape (batch_size, seq_len, d_model) or similar

# The forward method accesses:
# - x.size(1) - sequence length dimension
# - x.size(0) - batch size dimension
# - self.pe - the positional encoding buffer

# Create a small test input tensor
batch_size = 2
seq_len = 3
d_model = 4

test_input = torch.randn(batch_size, seq_len, d_model)

ARGS = {
    'd_model': d_model,
    'max_len': 10
}
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 PositionalEncoding — 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 PositionalEncoding 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
TemporalConv Ran
ran on a synthesized input · constructed an object; no result was computed
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("bb9b0666242adf98")
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, TemporalConv: 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)

# TemporalConv.__init__ parameters:
# - in_channels: int
# - out_channels: int
# - kernel_size: int
# - stride: int (default=1)
# - dilation: int (default=1)

ARGS = {
    "in_channels": 3,
    "out_channels": 4,
    "kernel_size": 5,
    "stride": 1,
    "dilation": 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 TemporalConv — 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 TemporalConv 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
TransformerEncoder Ran
ran on a synthesized input · constructed an object; no result was computed
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("5bbd4da441a48e72")
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, TransformerEncoder: 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)

class MockEncoderLayer(nn.Module):
    """Mock encoder layer that satisfies the interface expected by TransformerEncoder"""
    def __init__(self, d_model=4):
        super().__init__()
        self.d_model = d_model
        self.linear = nn.Linear(d_model, d_model)
    
    def forward(self, src, src_attn_mask=None, src_key_padding_mask=None, src_pos=None, verbose=False):
        # src shape: (seq_len, batch, d_model)
        output = self.linear(src)
        # Return output and a dummy attention map
        attn_map = torch.zeros(src.shape[1], src.shape[0], src.shape[0])
        return output, attn_map

class MockNorm(nn.Module):
    """Mock normalization layer"""
    def __init__(self, d_model=4):
        super().__init__()
        self.norm = nn.LayerNorm(d_model)
    
    def forward(self, x):
        return self.norm(x)

# Create encoder layer
encoder_layer = MockEncoderLayer(d_model=4)

# Number of layers
num_layers = 2

# Normalization layer
norm = MockNorm(d_model=4)

ARGS = {
    "encoder_layer": encoder_layer,
    "num_layers": num_layers,
    "norm": norm
}
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 TransformerEncoder — 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 TransformerEncoder 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
TransformerEncoderLayer Ran
ran on a synthesized input · constructed an object; no result was computed
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("f8f7be5c3f791c68")
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, TransformerEncoderLayer: 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
from torch import Tensor

torch.manual_seed(0)

# Create tiny tensors for testing
batch_size = 2
seq_len = 3
d_model = 4
nhead = 2
dim_feedforward = 8

# Source tensor: (seq_len, batch_size, d_model)
src = torch.randn(seq_len, batch_size, d_model)

# Optional positional embedding: (seq_len, batch_size, d_model)
src_pos = torch.randn(seq_len, batch_size, d_model)

# Optional attention mask: (seq_len, seq_len)
src_attn_mask = None

# Optional key padding mask: (batch_size, seq_len)
src_key_padding_mask = None

ARGS = {
    'd_model': d_model,
    'nhead': nhead,
    'dim_feedforward': dim_feedforward,
    'dropout': 0.1,
    'activation': 'relu',
    'normalize_before': False,
}
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 TransformerEncoderLayer — 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 TransformerEncoderLayer 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
Transformer_Encoder Ran
ran on a synthesized input · constructed an object; no result was computed
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("12e7bd98c7ee2907")
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, Transformer_Encoder: 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)

# Parameters for Transformer_Encoder
d_model = 4
nhead = 2
num_encoder_layers = 2
dim_feedforward = 8
dropout = 0.1
activation = 'relu'
normalize_before = False

ARGS = {
    'd_model': d_model,
    'nhead': nhead,
    'num_encoder_layers': num_encoder_layers,
    'dim_feedforward': dim_feedforward,
    'dropout': dropout,
    'activation': activation,
    'normalize_before': normalize_before,
}
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 Transformer_Encoder — 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 Transformer_Encoder 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_clones Ran
ran; our contract draft was wrong, not the code
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("df625bd27a59d895")
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 ModuleList. 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 copy
import math

torch.manual_seed(0)

# Create a simple nn.Module to use as the module parameter
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create instances for the arguments
module = SimpleModule()
N = 5

ARGS = {
    "module": module,
    "N": N
}
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 ModuleList — a ModuleList output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ModuleList 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
activation_factory Ran
ran; our contract draft was wrong, not the code
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("5426f077b824bbb4")
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 ReLU. 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)

ARGS = {
    'name': 'relu',
    'inplace': 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 ReLU — a ReLU output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ReLU 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
normalize_tensor Ran
ran, violated the contract we drafted
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("61e29634518807c2")
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, 4, 3, 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

torch.manual_seed(0)

# Create a tensor of shape (N, C, T, M) where:
# N=2 (batch size)
# C=4 (channel dimension to normalize over)
# T=3 (time dimension)
# M=5 (another dimension)
tensor = torch.randn(2, 4, 3, 5)

ARGS = {
    "tensor": tensor
}
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, 4, 3, 5) — shape (2, 4, 3, 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
HF_Pose Not yet run
unverified (harvested, no recorded run)
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("619c6b2598ece3bf")
How it ran

no invocation recorded.

MicroactionEncoder Not yet run
unverified (harvested, no recorded run)
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("0ed7906dcdd465df")
How it ran

no invocation recorded.

Trajectory_SelfAttention Not yet run
unverified (harvested, no recorded run)
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("e5491621abfadb10")
How it ran

no invocation recorded.

Unimodal_TF Not yet run
unverified (harvested, no recorded run)
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("b9201db3a21104d5")
How it ran

no invocation recorded.

fc_init Not yet run
unverified (harvested, no recorded run)
s-shamil/HandFormer/HandFormer/models/hf_pose.py
code served (permissive licence) · get_code("5870cc705fd8d62e")
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, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 1 the same month, 1 later) and 11 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
_get_clones
brick of this paper's assembly: helper
73 other papers
56 dated earlier, 1 the same month, 16 later · 59 only as an assembly brick
1706.03762 (2017-06)
Attention Is All You Need
2005.12872 (2020-05)
End-to-End Object Detection with Transformers
2010.04159 (2020-10)
Deformable DETR: Deformable Transformers for End-to-End Object Detection
and 70 more not named here
activation_factory
brick of this paper's assembly: helper
2 other papers
2 dated earlier
2208.05318 (2022-08)
Generative Action Description Prompts for Skeleton-based Action Recognition
2312.03288 (2023-12)
STEP CATFormer: Spatial-Temporal Effective Body-Part Cross Attention Transformer for Skeleton-based Action Recognition

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

3D hand pose is an underexplored modality for action recognition. Poses are compact yet informative and can greatly benefit applications with limited compute budgets. However, poses alone offer an incomplete understanding of actions, as they cannot fully capture objects and environments with which humans interact. We propose Hand-Former, a novel multimodal transformer, to efficiently model hand-object interactions. HandFormer combines 3D hand poses at a high temporal resolution for fine-grained motion modeling with sparsely sampled RGB frames for encoding scene semantics. Observing the unique characteristics of hand poses, we temporally factorize hand modeling and represent each joint by its short-term trajectories. This factorized pose representation combined with sparse RGB samples is remarkably efficient and highly accurate. Unimodal HandFormer with only hand poses outperforms existing skeleton-based methods at 5× fewer FLOPs. With RGB, we achieve new state-of-the-art performance on Assembly101 and H2O with significant improvements in egocentric action recognition.

For agents

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

get_harvested_code_for_paper("2403.09805")
get_code_for_paper("2403.09805")
have("2403.09805")

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

curl https://syntology.ai/api/ran/2403.09805.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.09805.svg)](https://syntology.ai/paper/2403.09805)

Connect an agent — have() is free.