SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.24037 · NeurIPS · 2024

TPC: Test-time Procrustes Calibration for Diffusion-based Human Image Animation

Chang Yoo, Sunjae Yoon, Gwanhyeong Koo, Younghwan Lee

arXiv · PDF · Open in the Atlas

Code that ran

10 of 14 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 0 have no contract verdict; 10 failures of our instrument; 4 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/2410.24037.json · badge: /api/ran/2410.24037.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
MooreThreads/Moore-AnimateAnyone canonical (classifier label)
classifier basis: census
9 ran · 4 no recorded run
dbstjswo505/TPC canonical (classifier label)
classifier basis not recorded on the link
1 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_snr Ran
ran; our contract draft was wrong, not the code
MooreThreads/Moore-AnimateAnyone/train_stage_1.py
code served (permissive licence) · get_code("16309f96889a07a1")
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 1, dtype family float, shape (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
import torch.nn as nn

torch.manual_seed(0)

class MockNoiseScheduler:
    """Minimal noise scheduler with alphas_cumprod attribute"""
    def __init__(self, num_timesteps=1000):
        # Create a simple decreasing schedule for alphas_cumprod
        # Shape: (num_timesteps,)
        self.alphas_cumprod = torch.linspace(1.0, 0.1, num_timesteps)

# Create noise scheduler
noise_scheduler = MockNoiseScheduler(num_timesteps=1000)

# Create timesteps tensor
# Shape: (2,) - batch of 2 timesteps
timesteps = torch.tensor([10, 50], dtype=torch.long)

ARGS = {
    "noise_scheduler": noise_scheduler,
    "timesteps": timesteps,
}
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 1, dtype family float, shape (2,) — shape (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
conv3x3 Ran
ran; our contract draft was wrong, not the code
dbstjswo505/TPC/TPC/densepose/modeling/hrnet.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.

decode_latents Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MooreThreads/Moore-AnimateAnyone/train_stage_2.py
code served (permissive licence) · get_code("2c02b3b459122f59")
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 5, dtype family float, shape (2, 3, 3, 5, 6) 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
from einops import rearrange

torch.manual_seed(0)

class MockVAEDecoder(nn.Module):
    """Minimal VAE decoder that returns a sample tensor."""
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
    
    def decode(self, latent):
        # latent shape: (1, 4, 3, 5) - (batch, channels, height, width)
        # output should be a sample with shape (1, 3, 3, 5) - (batch, 3 channels, height, width)
        batch_size = latent.shape[0]
        height = latent.shape[2]
        width = latent.shape[3]
        # Return a mock output with 3 channels (RGB-like)
        output = torch.randn(batch_size, 3, height, width)
        
        class DecodedOutput:
            def __init__(self, sample):
                self.sample = sample
        
        return DecodedOutput(output)

vae = MockVAEDecoder()

# Create latents tensor with shape (batch=2, channels=4, frames=3, height=5, width=6)
latents = torch.randn(2, 4, 3, 5, 6)

ARGS = {
    "vae": vae,
    "latents": latents,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 5, dtype family float, shape (2, 3, 3, 5, 6) — shape (2, 3, 3, 5, 6) 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
multiclass_nms Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MooreThreads/Moore-AnimateAnyone/src/dwpose/onnxdet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from BoyuanJiang/FitDiT, is pointer only (licence: NOASSERTION) · get_code("8ad410cc3c1eb4b2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from BoyuanJiang/FitDiT, and GitHub could not classify the licence of that repository (NOASSERTION). 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.

nms Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MooreThreads/Moore-AnimateAnyone/src/dwpose/onnxdet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from BoyuanJiang/FitDiT, is pointer only (licence: NOASSERTION) · get_code("ce701a4106957782")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from BoyuanJiang/FitDiT, and GitHub could not classify the licence of that repository (NOASSERTION). 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.

padRightDownCorner Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MooreThreads/Moore-AnimateAnyone/src/dwpose/util.py
code served (permissive licence) · get_code("3fc42c6bc0ad77b7")
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 test image with shape (height, width, channels)
# Using distinct dimensions: 5 (height), 7 (width), 3 (channels)
img = np.random.rand(5, 7, 3).astype(np.float32)

# stride parameter - controls padding alignment
stride = 4

# padValue - the value to use for padding
padValue = 0.0

ARGS = {
    "img": img,
    "stride": stride,
    "padValue": padValue
}
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
smart_resize Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
MooreThreads/Moore-AnimateAnyone/src/dwpose/util.py
code served (permissive licence) · get_code("a5bafc51aad44ec0")
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 int, shape (20, 24) 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 test images with different configurations

# Test case 1: 2D grayscale image (ndim == 2)
x_2d = np.random.randint(0, 256, size=(10, 12), dtype=np.uint8)

# Test case 2: 3D image with 1 channel
x_1channel = np.random.randint(0, 256, size=(10, 12, 1), dtype=np.uint8)

# Test case 3: 3D image with 3 channels (RGB)
x_3channel = np.random.randint(0, 256, size=(10, 12, 3), dtype=np.uint8)

# Test case 4: 3D image with multiple channels (not 1 or 3)
x_multichannel = np.random.randint(0, 256, size=(10, 12, 5), dtype=np.uint8)

# Target size
s = (20, 24)

# Use the 2D case as the primary test
ARGS = {
    'x': x_2d,
    's': s
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, 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 int, shape (20, 24) — shape (20, 24) 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
smart_resize_k Ran
ran; our contract draft was wrong, not the code
MooreThreads/Moore-AnimateAnyone/src/dwpose/util.py
code served (permissive licence) · get_code("8984b7a76d3499da")
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 (8, 22, 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 test image data
# 2D grayscale image
image_2d = np.random.randint(0, 256, size=(10, 15), dtype=np.uint8)

# 3D RGB image (1 channel)
image_1channel = np.random.randint(0, 256, size=(10, 15, 1), dtype=np.uint8)

# 3D RGB image (3 channels)
image_3channel = np.random.randint(0, 256, size=(10, 15, 3), dtype=np.uint8)

# 3D image with 4 channels (will use recursive stacking)
image_4channel = np.random.randint(0, 256, size=(10, 15, 4), dtype=np.uint8)

# Scaling factors
fx = 1.5  # horizontal scale factor
fy = 0.8  # vertical scale factor

# Test with 2D image and scale factors
ARGS = {
    "x": image_3channel,
    "fx": fx,
    "fy": fy
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, 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 (8, 22, 3) — shape (8, 22, 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
torch_dfs Ran
ran; our contract draft was wrong, not the code
MooreThreads/Moore-AnimateAnyone/src/models/mutual_self_attention.py
code served (permissive licence) · get_code("e55f24470648724c")
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 torch.nn as nn

torch.manual_seed(0)

# Build a small model with a hierarchical structure
# The function calls model.children(), so we need a model with child modules
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.layer1 = nn.Linear(4, 3)
        self.layer2 = nn.Sequential(
            nn.Linear(3, 2),
            nn.ReLU()
        )
        self.layer3 = nn.Linear(2, 1)
    
    def forward(self, x):
        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        return x

model = SimpleModel()

ARGS = {
    "model": model
}
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 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
zero_module Ran
ran; our contract draft was wrong, not the code
MooreThreads/Moore-AnimateAnyone/src/models/motion_module.py
code served (permissive licence) · get_code("4719c763c53be3fe")
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 SimpleModule. 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 a simple nn.Module with some parameters
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
        self.conv = nn.Conv1d(2, 5, 3)
    
    def forward(self, x):
        return x

module = SimpleModule()

ARGS = {
    "module": module
}
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 SimpleModule — a SimpleModule output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModule 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
demo_postprocess Not yet run
unverified (harvested, no recorded run)
MooreThreads/Moore-AnimateAnyone/src/dwpose/onnxdet.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from BoyuanJiang/FitDiT, is pointer only (licence: NOASSERTION) · get_code("64d861666d61123b")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from BoyuanJiang/FitDiT, and GitHub could not classify the licence of that repository (NOASSERTION). 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. No invocation recorded.

get_motion_module Not yet run
unverified (harvested, no recorded run)
MooreThreads/Moore-AnimateAnyone/src/models/motion_module.py
code served (permissive licence) · get_code("525d02f457ec7c07")
How it ran

no invocation recorded.

postprocess Not yet run
unverified (harvested, no recorded run)
MooreThreads/Moore-AnimateAnyone/src/dwpose/onnxpose.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from lihxxx/DisPose, is pointer only (licence: NONE) · get_code("bb3bf2f13b8a3eb5")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from lihxxx/DisPose, 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. No invocation recorded.

preprocess Not yet run
unverified (harvested, no recorded run)
MooreThreads/Moore-AnimateAnyone/src/dwpose/onnxpose.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from lihxxx/DisPose, is pointer only (licence: NONE) · get_code("bfd83ef0068198d1")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from lihxxx/DisPose, 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. No invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 14 are also present in other papers whose code we hold (14 with at least one paper dated earlier, 2 the same month, 10 later) and 0 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
629 dated earlier, 6 the same month, 34 later, 17 same year, order not determinable · 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
zero_module 196 other papers
149 dated earlier, 6 the same month, 40 later, 1 same year, order not determinable · 49 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2105.05233 (2021-05)
Diffusion Models Beat GANs on Image Synthesis
and 193 more not named here
get_motion_module 11 other papers
10 dated earlier, 1 later
2310.19512 (2023-10)
VideoCrafter1: Open Diffusion Models for High-Quality Video Generation
2311.00213 (2023-11)
Consistent Video-to-Video Transfer Using Synthetic Dataset
2312.13964 (2023-12)
PIA: Your Personalized Image Animator via Plug-and-Play Modules in Text-to-Image Models
and 8 more not named here
compute_snr 9 other papers
5 dated earlier, 3 later, 1 undated
2311.17918 (2023-11)
Driving into the Future: Multiview Visual Forecasting and Planning with World Model for Autonomous Driving
2403.10094 (2024-03)
RangeLDM: Fast Realistic LiDAR Point Cloud Generation
2403.12931 (2024-03)
You Only Sample Once: Taming One-Step Text-To-Image Synthesis by Self-Cooperative Diffusion GANs
and 6 more not named here
torch_dfs 9 other papers
7 dated earlier, 2 later
2403.14781 (2024-03)
Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance
2403.17694 (2024-03)
AniPortrait: Audio-Driven Synthesis of Photorealistic Portrait Animation
2405.11794 (2024-05)
ViViD: Video Virtual Try-on using Diffusion Models
and 6 more not named here
demo_postprocess 3 other papers
1 dated earlier, 2 later
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
2411.10499 (2024-11)
FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try-on
2412.09349 (2024-12)
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
multiclass_nms 3 other papers
1 dated earlier, 2 later
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
2411.10499 (2024-11)
FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try-on
2412.09349 (2024-12)
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
nms 3 other papers
1 dated earlier, 2 later
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
2411.10499 (2024-11)
FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try-on
2412.09349 (2024-12)
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
padRightDownCorner 2 other papers
2 dated earlier
1611.08050 (2016-11)
Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
postprocess 2 other papers
1 dated earlier, 1 later
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
2412.09349 (2024-12)
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
preprocess 2 other papers
1 dated earlier, 1 later
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
2412.09349 (2024-12)
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
decode_latents 1 other paper
1 dated earlier
2403.17694 (2024-03)
AniPortrait: Audio-Driven Synthesis of Photorealistic Portrait Animation
smart_resize 1 other paper
1 dated earlier
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation
smart_resize_k 1 other paper
1 dated earlier
2307.15880 (2023-07)
Effective Whole-body Pose Estimation with Two-stages Distillation

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

Human image animation aims to generate a human motion video from the inputs of a reference human image and a target motion video. Current diffusion-based image animation systems exhibit high precision in transferring human identity into targeted motion, yet they still exhibit irregular quality in their outputs. Their optimal precision is achieved only when the physical compositions (i.e., scale and rotation) of the human shapes in the reference image and target pose frame are aligned. In the absence of such alignment, there is a noticeable decline in fidelity and consistency. Especially, in real-world environments, this compositional misalignment commonly occurs, posing significant challenges to the practical usage of current systems. To this end, we propose Test-time Procrustes Calibration (TPC), which enhances the robustness of diffusion-based image animation systems by maintaining optimal performance even when faced with compositional misalignment, effectively addressing real-world scenarios. The TPC provides a calibrated reference image for the diffusion model, enhancing its capability to understand the correspondence between human shapes in the reference and target images. Our method is simple and can be applied to any diffusion-based image animation system in a model-agnostic manner, improving the effectiveness at test time without additional training. The project is available at github.com/dbstjswo505/TPC

For agents

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

get_harvested_code_for_paper("2410.24037")
get_code_for_paper("2410.24037")
have("2410.24037")

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

curl https://syntology.ai/api/ran/2410.24037.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.24037.svg)](https://syntology.ai/paper/2410.24037)

Connect an agent — have() is free.