SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.09116 · CVPR · 2022

MotionAug: Augmentation with Physical Correction for Human Motion Prediction

Takahiro Maeda

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 (0 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; 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/2203.09116.json · badge: /api/ran/2203.09116.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
meaten/MotionAug community
Papers with Code archive
5 ran · 0 no recorded run
FunctionStatusWhere it lives
Estimator_length Ran
ran on a synthesized input
meaten/MotionAug/motionAE/src/models.py
code served (permissive licence) · get_code("af8ae61e28c73e64")
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 Tensor, ndim 2, dtype family float, shape (2, 1) 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

class Estimator_length(nn.Module):

    def __init__(self, dim_z):
        super(Estimator_length, self).__init__()
        self.dim_z = dim_z
        self.linear1 = nn.Linear(in_features=self.dim_z, out_features=int(self.dim_z / 4))
        self.linear2 = nn.Linear(in_features=int(self.dim_z / 4), out_features=1)
        self.activation = nn.ReLU()

    def forward(self, x):
        x = self.activation(self.linear1(x))
        return self.linear2(x)


INIT_ARGS = {
    'dim_z': 8
}

ARGS = {
    'x': torch.randn(2, 8)
}

if __name__ == '__main__':
    model = Estimator_length(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Output: {output}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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, 1) — shape (2, 1) 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
lstmDecoder Ran
ran on a synthesized input
meaten/MotionAug/motionAE/src/models.py
code served (permissive licence) · get_code("db79e6b023c85ab7")
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 Tensor, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 numpy as np
from scipy import stats

INIT_ARGS = {
    "input_len": 3,
    "dim": 4,
    "dim_z": 5,
    "dropout": 0.1,
}

# Construct input tensor for forward pass
batch_size = 2
x = torch.randn(batch_size, INIT_ARGS["dim_z"])

ARGS = {
    "x": x,
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
numpy, scipy, 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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
lstmEncoder_vae Ran
ran on a synthesized input
meaten/MotionAug/motionAE/src/models.py
code served (permissive licence) · get_code("fce9adf35d72f410")
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 list. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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

class lstmEncoder_vae(nn.Module):

    def __init__(self, input_len, dim, dim_z, dropout=0.0):
        super(lstmEncoder_vae, self).__init__()
        self.input_len = input_len
        self.dim = dim
        self.dim_z = dim_z
        self.linear = nn.Linear(in_features=self.dim, out_features=self.dim_z)
        self.norm = nn.LayerNorm([self.input_len, self.dim])
        self.rnn = nn.LSTM(input_size=self.dim_z, hidden_size=self.dim_z * 2, num_layers=1, batch_first=True, dropout=dropout, bidirectional=True)
        self.mu_linear1 = nn.Linear(in_features=self.dim_z * 4, out_features=self.dim_z * 2)
        self.mu_linear2 = nn.Linear(in_features=self.dim_z * 2, out_features=self.dim_z)
        self.log_var_linear1 = nn.Linear(in_features=self.dim_z * 4, out_features=self.dim_z * 2)
        self.log_var_linear2 = nn.Linear(in_features=self.dim_z * 2, out_features=self.dim_z)
        self.activation = nn.ReLU()

    def forward(self, x, x_lengths):
        x = self.norm(x)
        x = self.activation(self.linear(x))
        x = torch.nn.utils.rnn.pack_padded_sequence(x, x_lengths, batch_first=True, enforce_sorted=False)
        out, (_, _) = self.rnn(x)
        out, out_lengths = torch.nn.utils.rnn.pad_packed_sequence(out, batch_first=True)
        out = torch.sum(out, dim=1) / out_lengths[:, None].type(torch.FloatTensor)
        mu = self.activation(self.mu_linear1(out))
        mu = self.mu_linear2(mu)
        log_var = self.activation(self.log_var_linear1(out))
        log_var = self.log_var_linear2(log_var)
        return [mu, log_var]


INIT_ARGS = {
    'input_len': 5,
    'dim': 8,
    'dim_z': 4,
    'dropout': 0.0
}

ARGS = {
    'x': torch.randn(2, 5, 8),
    'x_lengths': torch.tensor([5, 3], dtype=torch.long)
}

if __name__ == '__main__':
    model = lstmEncoder_vae(**INIT_ARGS)
    model.eval()
    output = model(**ARGS)
    print("Output shapes:")
    print(f"  mu shape: {output[0].shape}")
    print(f"  log_var shape: {output[1].shape}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
lstmVAE Ran
ran on a synthesized input
meaten/MotionAug/motionAE/src/models.py
code served (permissive licence) · get_code("cc1b52c9542d2a0e")
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 list. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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

# ============================================================================
# CLASS DEFINITIONS (from the problem statement)
# ============================================================================

class Estimator_length(nn.Module):

    def __init__(self, dim_z):
        super(Estimator_length, self).__init__()
        self.dim_z = dim_z
        self.linear1 = nn.Linear(in_features=self.dim_z, out_features=int(self.dim_z / 4))
        self.linear2 = nn.Linear(in_features=int(self.dim_z / 4), out_features=1)
        self.activation = nn.ReLU()

    def forward(self, x):
        x = self.activation(self.linear1(x))
        return self.linear2(x)

class lstmDecoder(nn.Module):

    def __init__(self, input_len, dim, dim_z, dropout=0.0):
        super(lstmDecoder, self).__init__()
        self.input_len = input_len
        self.dim = dim
        self.dim_z = dim_z
        self.linear = nn.Linear(in_features=self.dim_z, out_features=self.dim_z)
        self.rnn = nn.LSTM(input_size=self.dim_z, hidden_size=self.dim_z * 2, num_layers=1, batch_first=True, dropout=dropout, bidirectional=False)
        self.linear1 = nn.Linear(in_features=self.dim_z * 2, out_features=self.dim_z)
        self.linear2 = nn.Linear(in_features=self.dim_z, out_features=self.dim)
        self.activation = nn.ReLU()

    def forward(self, x):
        x = self.activation(self.linear(x))
        x = x[:, None, :].expand(-1, self.input_len, -1)
        out, (_, _) = self.rnn(x)
        out = self.activation(self.linear1(out))
        recon = self.linear2(out)
        return recon

class lstmEncoder_vae(nn.Module):

    def __init__(self, input_len, dim, dim_z, dropout=0.0):
        super(lstmEncoder_vae, self).__init__()
        self.input_len = input_len
        self.dim = dim
        self.dim_z = dim_z
        self.linear = nn.Linear(in_features=self.dim, out_features=self.dim_z)
        self.norm = nn.LayerNorm([self.input_len, self.dim])
        self.rnn = nn.LSTM(input_size=self.dim_z, hidden_size=self.dim_z * 2, num_layers=1, batch_first=True, dropout=dropout, bidirectional=True)
        self.mu_linear1 = nn.Linear(in_features=self.dim_z * 4, out_features=self.dim_z * 2)
        self.mu_linear2 = nn.Linear(in_features=self.dim_z * 2, out_features=self.dim_z)
        self.log_var_linear1 = nn.Linear(in_features=self.dim_z * 4, out_features=self.dim_z * 2)
        self.log_var_linear2 = nn.Linear(in_features=self.dim_z * 2, out_features=self.dim_z)
        self.activation = nn.ReLU()

    def forward(self, x, x_lengths):
        x = self.norm(x)
        x = self.activation(self.linear(x))
        x = torch.nn.utils.rnn.pack_padded_sequence(x, x_lengths, batch_first=True, enforce_sorted=False)
        out, (_, _) = self.rnn(x)
        out, out_lengths = torch.nn.utils.rnn.pad_packed_sequence(out, batch_first=True)
        out = torch.sum(out, dim=1) / out_lengths[:, None].type(torch.FloatTensor)
        mu = self.activation(self.mu_linear1(out))
        mu = self.mu_linear2(mu)
        log_var = self.activation(self.log_var_linear1(out))
        log_var = self.log_var_linear2(log_var)
        return [mu, log_var]

def reparametrize(mu, log_var):
    std = torch.exp(0.5 * log_var)
    eps = torch.randn_like(mu)
    return eps * std + mu

class lstmVAE(nn.Module):

    def __init__(self, input_len, dim, dim_z):
        super(lstmVAE, self).__init__()
        self.encoder = lstmEncoder_vae(input_len, dim, dim_z)
        self.decoder = lstmDecoder(input_len, dim, dim_z)
        self.estimator_length = Estimator_length(dim_z)

    def forward(self, x, x_lengths):
        mu, log_var = self.encoder(x, x_lengths)
        z = reparametrize(mu, log_var)
        return [x, self.decoder(z), x_lengths, mu, log_var, z, self.estimator_length(z)]

# ============================================================================
# SETUP CODE
# ============================================================================

INIT_ARGS = {
    'input_len': 5,
    'dim': 3,
    'dim_z': 4,
}

ARGS = {
    'x': torch.randn(2, 5, 3),
    'x_lengths': torch.tensor([5, 4], dtype=torch.long),
}

if __name__ == '__main__':
    model = lstmVAE(**INIT_ARGS)
    model.eval()
    output = model(**ARGS)
    print("Output structure:")
    print(f"  x: {output[0].shape}")
    print(f"  decoder output: {output[1].shape}")
    print(f"  x_lengths: {output[2]}")
    print(f"  mu: {output[3].shape}")
    print(f"  log_var: {output[4].shape}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
reparametrize Ran
ran; our contract draft was wrong, not the code
meaten/MotionAug/motionAE/src/models.py
code served (permissive licence) · get_code("0dac1f8de6997327")
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 small tensors for mu and log_var
# mu: mean of the distribution, shape (batch_size, latent_dim)
# log_var: log variance of the distribution, shape (batch_size, latent_dim)
mu = torch.randn(2, 4)
log_var = torch.randn(2, 4)

ARGS = {
    "mu": mu,
    "log_var": log_var,
}
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

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, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 5 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.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

This paper presents a motion data augmentation scheme incorporating motion synthesis encouraging diversity and motion correction imposing physical plausibility. This motion synthesis consists of our modified Variational AutoEncoder (VAE) and Inverse Kinematics (IK). In this VAE, our proposed sampling-nearsamples method generates various valid motions even with insufficient training motion data. Our IK-based motion synthesis method allows us to generate a variety of motions semi-automatically. Since these two schemes generate unrealistic artifacts in the synthesized motions, our motion correction rectifies them. This motion correction scheme consists of imitation learning with physics simulation and subsequent motion debiasing. For this imitation learning, we propose the PD-residual force that significantly accelerates the training process. Furthermore, our motion debiasing successfully offsets the motion bias induced by imitation learning to maximize the effect of augmentation. As a result, our method outperforms previous noise-based motion augmentation methods by a large margin on both Recurrent Neural Network-based and Graph Convolutional Network-based human motion prediction models. The code is available at https: //github.com/meaten/MotionAug.

For agents

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

get_harvested_code_for_paper("2203.09116")
get_code_for_paper("2203.09116")
have("2203.09116")

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

curl https://syntology.ai/api/ran/2203.09116.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.09116.svg)](https://syntology.ai/paper/2203.09116)

Connect an agent — have() is free.