SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2307.07754 · ICCV · 2023

Bidirectionally Deformable Motion Modulation For Video-based Human Pose Transfer

Ray Cheung, Kun Li, Yu Xue, Lai-Man Po, Yuzhi Zhao, Wing-Yin Yu

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (4 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 1 failure of our instrument; 0 no recorded run), measured 2026-11-22; 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/2307.07754.json · badge: /api/ran/2307.07754.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
rocketappslab/bdmm marked official in Papers with Code
Papers with Code archive
5 ran · 0 no recorded run
FunctionStatusWhere it lives
EqualLinear Ran
ran on a synthesized input · constructed an object; no result was computed
rocketappslab/bdmm/model/networks/ModulatedDCN.py
code served (permissive licence) · get_code("760797bc84c2e0b6")
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, EqualLinear: 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

torch.manual_seed(0)

ARGS = {
    "in_dim": 4,
    "out_dim": 3,
    "bias": True,
    "bias_init": 0,
    "lr_mul": 1,
    "activation": None,
}
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 EqualLinear — 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 EqualLinear 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
FusedLeakyReLUFunction Ran
ran on a synthesized input · constructed an object; no result was computed
rocketappslab/bdmm/model/networks/ModulatedDCN.py
code served (permissive licence) · get_code("b232065a48983f60")
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, FusedLeakyReLUFunction: 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)

# Create mock fused module since it's not available in standard PyTorch
class MockFused:
    @staticmethod
    def fused_bias_act(input_tensor, bias, out_tensor, act, grad, negative_slope, scale):
        # Mock implementation that returns a tensor of the same shape as input
        result = input_tensor.clone()
        if bias is not None and bias.numel() > 0:
            result = result + bias.view(-1, 1) if result.ndim > 1 else result + bias
        result = result * scale
        # Apply leaky relu if needed
        result = torch.where(result > 0, result, result * negative_slope)
        return result

# Mock the fused module
import sys
from types import ModuleType
fused = ModuleType('fused')
fused.fused_bias_act = MockFused.fused_bias_act
sys.modules['fused'] = fused

# Now we can define the arguments
# Parameters for FusedLeakyReLUFunction.forward:
# - input: tensor
# - bias: tensor or None
# - negative_slope: float
# - scale: float

input_tensor = torch.randn(2, 4)
bias_tensor = torch.randn(4)
negative_slope = 0.2
scale = 1.0

ARGS = {
    'input': input_tensor,
    'bias': bias_tensor,
    'negative_slope': negative_slope,
    'scale': scale
}
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 FusedLeakyReLUFunction — 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 FusedLeakyReLUFunction 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
FusedLeakyReLUFunctionBackward Ran
ran on a synthesized input · constructed an object; no result was computed
rocketappslab/bdmm/model/networks/ModulatedDCN.py
code served (permissive licence) · get_code("127d346428e96ec6")
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, FusedLeakyReLUFunctionBackward: 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)

# Mock the 'fused' module since it's an external dependency
class MockFused:
    @staticmethod
    def fused_bias_act(input_tensor, bias, out, act, dim, negative_slope, scale):
        # Simple mock implementation that returns a tensor of same shape as input
        return input_tensor * scale

fused = MockFused()

# Create small test tensors
batch_size = 2
feature_dim = 4
spatial_dim = 3

grad_output = torch.randn(batch_size, feature_dim, spatial_dim, spatial_dim)
out = torch.randn(batch_size, feature_dim, spatial_dim, spatial_dim)
bias = True
negative_slope = 0.2
scale = 1.0

ARGS = {
    "grad_output": grad_output,
    "out": out,
    "bias": bias,
    "negative_slope": negative_slope,
    "scale": scale,
}
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 FusedLeakyReLUFunctionBackward — 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 FusedLeakyReLUFunctionBackward 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
ModulatedDCN2dv2 Ran
ran on a synthesized input · constructed an object; no result was computed
rocketappslab/bdmm/model/networks/ModulatedDCN.py
code served (permissive licence) · get_code("e09bee27d7c071ee")
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, ModulatedDCN2dv2: 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

torch.manual_seed(0)

# Create a small stand-in for the `fused` module that's imported but not shown
class FusedModule:
    @staticmethod
    def fused_bias_act(input, bias, out, act, grad, negative_slope, scale):
        # Simple stand-in: just apply leaky relu
        if grad == 0:  # forward
            if bias.numel() > 0:
                rest_dim = [1] * (input.ndim - bias.ndim - 1)
                result = input + bias.view(1, bias.shape[0], *rest_dim)
            else:
                result = input
            return torch.nn.functional.leaky_relu(result, negative_slope=negative_slope) * scale
        else:  # backward
            return torch.nn.functional.leaky_relu(input, negative_slope=negative_slope) * scale

# Monkey-patch the fused module into the global namespace for the forward pass
import sys
fused_module = FusedModule()
sys.modules['fused'] = type(sys)('fused')
sys.modules['fused'].fused_bias_act = FusedModule.fused_bias_act

# Now build the arguments
batch_size = 2
in_channel = 4
out_channel = 3
kernel_size = 3
style_dim = 5
height = 8
width = 8

input_tensor = torch.randn(batch_size, in_channel, height, width)
style_tensor = torch.randn(batch_size, style_dim)
offset_tensor = torch.randn(batch_size, 2 * kernel_size * kernel_size, height, width)
mask_tensor = torch.randn(batch_size, kernel_size * kernel_size, height, width)

ARGS = {
    'in_channel': in_channel,
    'out_channel': out_channel,
    'kernel_size': kernel_size,
    'style_dim': style_dim,
    'demodulate': True,
}

# Create an instance to test that initialization works
module = nn.Module()
for key, val in ARGS.items():
    setattr(module, f'_arg_{key}', val)
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ModulatedDCN2dv2 — 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 ModulatedDCN2dv2 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
fused_leaky_relu Ran
ran; our contract draft was wrong, not the code
rocketappslab/bdmm/model/networks/ModulatedDCN.py
code served (permissive licence) · get_code("10b5ed33405afefb")
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, 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
import torch.nn as nn

torch.manual_seed(0)

# Create minimal test tensors
# input: batch_size=2, channels=4, spatial=3x3
input_tensor = torch.randn(2, 4, 3, 3, requires_grad=True)

# bias: channels=4
bias_tensor = torch.randn(4, requires_grad=True)

# Parameters
negative_slope = 0.2
scale = 2 ** 0.5

ARGS = {
    'input': input_tensor,
    'bias': bias_tensor,
    'negative_slope': negative_slope,
    'scale': scale,
}
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, 3) — shape (2, 4, 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

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 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
EqualLinear
brick of this paper's assembly: helper
8 other papers
8 dated earlier
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
2101.02477 (2021-01)
GAN-Control: Explicitly Controllable GANs
2103.17249 (2021-03)
StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery
and 5 more not named here

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

Video-based human pose transfer is a video-to-video generation task that animates a plain source human image based on a series of target human poses. Considering the difficulties in transferring highly structural patterns on the garments and discontinuous poses, existing methods often generate unsatisfactory results such as distorted textures and flickering artifacts. To address these issues, we propose a novel Deformable Motion Modulation (DMM) that utilizes geometric kernel offset with adaptive weight modulation to simultaneously perform feature alignment and style transfer. Different from normal style modulation used in style transfer, the proposed modulation mechanism adaptively reconstructs smoothed frames from style codes according to the object shape through an irregular receptive field of view. To enhance the spatio-temporal consistency, we leverage bidirectional propagation to extract the hidden motion information from a warped image sequence generated by noisy poses. The proposed feature propagation significantly enhances the motion prediction ability by forward and backward propagation. Both quantitative and qualitative experimental results demonstrate superiority over the state-of-the-arts in terms of image fidelity and visual continuity. The source code is publicly available at github.com/rocketappslab/bdmm.

For agents

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

get_harvested_code_for_paper("2307.07754")
get_code_for_paper("2307.07754")
have("2307.07754")

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

curl https://syntology.ai/api/ran/2307.07754.json

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

[![Syntology run record](https://syntology.ai/api/ran/2307.07754.svg)](https://syntology.ai/paper/2307.07754)

Connect an agent — have() is free.