SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2407.16124 · 2024

Fréchet Video Motion Distance: A Metric for Evaluating Motion Consistency in Videos

arXiv · PDF · Open in the Atlas

Code that ran

20 of 23 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; 8 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 11 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/2407.16124.json · badge: /api/ran/2407.16124.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
ljh0v0/fmd-frechet-motion-distance canonical (classifier label)
classifier basis: census
18 ran · 2 no recorded run
dsl-lab/fvmd-frechet-video-motion-distance community (classifier label)
classifier basis: plan_row
2 ran · 1 no recorded run
FunctionStatusWhere it lives
bilinear_sample2d Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/samp.py
code served (permissive licence) · get_code("a1c2b83f2ed58af9")
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 torch

torch.manual_seed(0)

# Construct arguments for bilinear_sample2d
# im: image tensor of shape (B, C, H, W)
# x: x coordinates of shape (B, N)
# y: y coordinates of shape (B, N)
# return_inbounds: boolean flag

B = 2  # batch size
C = 4  # channels
H = 8  # height
W = 6  # width
N = 5  # number of sample points

# Create a small image tensor (B, C, H, W)
im = torch.randn(B, C, H, W)

# Create coordinate tensors (B, N) with values in valid range
# x coordinates should be in [0, W-1], y coordinates in [0, H-1]
x = torch.tensor([
    [1.5, 2.3, 3.7, 4.2, 5.0],
    [0.5, 1.2, 2.8, 3.5, 4.9]
], dtype=torch.float32)

y = torch.tensor([
    [1.2, 2.5, 3.1, 4.8, 5.5],
    [0.8, 1.9, 2.7, 3.3, 6.2]
], dtype=torch.float32)

# Flag for returning inbounds mask
return_inbounds = True

ARGS = {
    'im': im,
    'x': x,
    'y': y,
    'return_inbounds': return_inbounds
}
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 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
calc_acceleration Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/keypoint_tracking.py
code served (permissive licence) · get_code("a6571e1c41b062fb")
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, 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)

# Construct a velocity tensor with shape (Batch_size, Frames, Points, 2)
# Using distinct sizes: B=2, S=4, N=3, last dim=2
velocity = torch.randn(2, 4, 3, 2)

ARGS = {
    "velocity": velocity
}
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, 2) — shape (2, 4, 3, 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
calc_velocity Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/keypoint_tracking.py
code served (permissive licence) · get_code("a6307e11ed350db2")
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, 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 a small trajectory tensor with shape (Batch_size, Frames, Points, 2)
# Using distinct sizes: batch=2, frames=4, points=3, coordinates=2
trajs_e = torch.randn(2, 4, 3, 2)

ARGS = {
    "trajs_e": trajs_e,
}
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, 2) — shape (2, 4, 3, 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
calculate_activation_statistics Ran
ran; our contract draft was wrong, not the code
ljh0v0/fmd-frechet-motion-distance/fvmd/frechet_distance.py
code served (permissive licence) · get_code("be7a68ab5541a79d")
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
import torch

torch.manual_seed(0)

# The function takes a numpy array v and computes statistics
# Looking at the body:
# - v.shape[0] is accessed (batch dimension)
# - v.reshape(B, -1) flattens all dimensions except batch
# - np.mean and np.cov are called on the flattened array

# Create a small test array: batch_size=2, with some spatial/feature dimensions
# Using shape (2, 4, 3) to have distinct dimensions
v = np.random.randn(2, 4, 3).astype(np.float32)

ARGS = {
    "v": v
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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: none by the harness; any seed is in the setup block
calculate_fd_given_vectors Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/frechet_distance.py
code served (permissive licence) · get_code("96b374030dd1b41b")
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 float64, ndim 0, 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 numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Create small feature vectors for two distributions
# feat1: batch_size=2, feature_dim=4
feat1 = np.random.randn(2, 4).astype(np.float32)

# feat2: batch_size=2, feature_dim=4
feat2 = np.random.randn(2, 4).astype(np.float32)

ARGS = {
    "feat1": feat1,
    "feat2": feat2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
calculate_frechet_distance Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/frechet_distance.py
code served (permissive licence) · get_code("1640133300f0b805")
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 float64, ndim 0, 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 numpy as np
from scipy import linalg

# Set random seed for reproducibility
np.random.seed(0)

# Create small sample data for Frechet Distance calculation
# mu1 and mu2 are mean vectors (1D arrays)
# sigma1 and sigma2 are covariance matrices (2D arrays)

dim = 4  # dimension of the feature space

# Generate mean vectors
mu1 = np.random.randn(dim)
mu2 = np.random.randn(dim)

# Generate positive semi-definite covariance matrices
# Create random matrices and multiply by their transpose to ensure positive semi-definiteness
A1 = np.random.randn(dim, dim)
sigma1 = A1 @ A1.T

A2 = np.random.randn(dim, dim)
sigma2 = A2 @ A2.T

# eps parameter with default value
eps = 1e-05

ARGS = {
    "mu1": mu1,
    "sigma1": sigma1,
    "mu2": mu2,
    "sigma2": sigma2,
    "eps": eps,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
coords_grid Ran
ran, honoured the contract we drafted
dsl-lab/fvmd-frechet-video-motion-distance/fvmd/nets/pips2.py
code served (permissive licence) · get_code("ea43be6c224959b1")
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 4, dtype family float, shape (2, 2, 4, 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)

ARGS = {
    "batch": 2,
    "ht": 4,
    "wd": 3,
}
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 4, dtype family float, shape (2, 2, 4, 3) — shape (2, 2, 4, 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
count_subcube_hist Ran
ran; our contract draft was wrong, not the code
ljh0v0/fmd-frechet-motion-distance/fvmd/extract_motion_features.py
code served (permissive licence) · get_code("0b86b901d7c8eee5")
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 1, dtype family float, shape (8,) 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

# Create a small vector_cell array with shape (S, H, W, 2)
# S=2, H=3, W=4 (distinct sizes for different dimensions)
S, H, W = 2, 3, 4
vector_cell = np.random.randn(S, H, W, 2).astype(np.float32)

# Define arguments
ARGS = {
    "vector_cell": vector_cell,
    "angle_bins": 8,
    "magnitude_bins": 256,
}
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 ndarray, ndim 1, dtype family float, shape (8,) — shape (8,) 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
cut_subcube Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/extract_motion_features.py
code served (permissive licence) · get_code("564904fe8c3e3f36")
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 video tensor with shape (B, S, H, W, 2)
# B=2 (batch), S=8 (frames), H=10 (height), W=12 (width), 2 (2D vectors)
vectors = np.random.randn(2, 8, 10, 12, 2).astype(np.float32)

ARGS = {
    "vectors": vectors,
    "cell_size": 5,
    "cube_frames": 4,
}
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
eye_3x3 Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/geom.py
code served (permissive licence) · get_code("c6a46e4312ecdb81")
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. 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, 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)

ARGS = {
    'B': 2,
    'device': 'cpu'
}
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
farthest_point_sample Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/misc.py
code served (permissive licence) · get_code("e1adde1a35ea6aa9")
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 int, shape (2, 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)

# Construct arguments for farthest_point_sample
xyz = torch.randn(2, 4, 3)  # B=2, N=4, C=3 (pointcloud data)
npoint = 2  # number of samples to select
include_ends = False
deterministic = False

ARGS = {
    "xyz": xyz,
    "npoint": npoint,
    "include_ends": include_ends,
    "deterministic": deterministic,
}
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 int, shape (2, 2) — shape (2, 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
get_lr_str Ran
ran; our contract draft was wrong, not the code
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/basic.py
code served (permissive licence) · get_code("1ef7e186129f3877")
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 str. 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 numpy as np

def get_lr_str(lr):
    lrn = '%.1e' % lr
    lrn = lrn[0] + lrn[3:5] + lrn[-1]
    return lrn

# The function takes a learning rate (float) and formats it as a string
# It expects a numeric value that can be formatted with %.1e

ARGS = {
    'lr': 0.001
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
ind2sub Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/basic.py
code served (permissive licence) · get_code("433a4096809820d8")
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, ndim 1, dtype family int; 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

torch.manual_seed(0)

ARGS = {
    "height": 10,
    "width": 8,
    "ind": 42
}
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 tuple, ndim 1, dtype family int — a tuple output has no array shape of its own; none is stated
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
load Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/saverloader.py
code served (permissive licence) · get_code("7cc0ed613ecce873")
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 int, ndim 0, dtype family int, shape () 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
import tempfile
import os

torch.manual_seed(0)

# Create a minimal model
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create a minimal optimizer
model = SimpleModel()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)

# Create a minimal scheduler
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=1)

# Create a temporary checkpoint directory with a valid checkpoint file
ckpt_dir = tempfile.mkdtemp()

# Create a checkpoint file
checkpoint = {
    'model_state_dict': model.state_dict(),
    'optimizer_state_dict': optimizer.state_dict(),
    'scheduler_state_dict': scheduler.state_dict()
}
ckpt_path = os.path.join(ckpt_dir, 'model-000000001.pth')
torch.save(checkpoint, ckpt_path)

# Define arguments
ARGS = {
    'ckpt_dir': ckpt_dir,
    'model': model,
    'optimizer': optimizer,
    'scheduler': scheduler,
    'step': 0,
    'model_name': 'model',
    'ignore_load': 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 int, ndim 0, dtype family int, shape () — shape () 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
matmul2 Ran
ran, honoured the contract we drafted
dsl-lab/fvmd-frechet-video-motion-distance/fvmd/utils/geom.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from ganwanshui/simpleoccupancy, is pointer only (licence: NONE) · get_code("c9c44e261dffcaf1")
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 ganwanshui/simpleoccupancy, 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.

matmul3 Ran
ran; our contract draft was wrong, not the code
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/geom.py
code served (permissive licence) · get_code("0d3e6ecf8bc2abdd")
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, 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 three small matrices for chained matrix multiplication
# Using distinct dimensions: 2x3, 3x4, 4x5
mat1 = torch.randn(2, 3)
mat2 = torch.randn(3, 4)
mat3 = torch.randn(4, 5)

ARGS = {
    "mat1": mat1,
    "mat2": mat2,
    "mat3": mat3,
}
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, 5) — shape (2, 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
posemb_sincos_2d_xy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/misc.py
code served (permissive licence) · get_code("1ef23d4ecc09d434")
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, 8) 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)

# xy: shape (B, S, D) where D=2, containing 2D coordinates
# C: feature dimension for sincos embedding (must be multiple of 4)
# temperature: scaling factor for the sinusoidal frequencies
# dtype: output data type
# cat_coords: whether to concatenate original coordinates to embedding

B = 2  # batch size
S = 3  # sequence length
D = 2  # coordinate dimension (always 2)
C = 8  # feature dimension (multiple of 4)

xy = torch.randn(B, S, D, dtype=torch.float32)

ARGS = {
    "xy": xy,
    "C": C,
    "temperature": 10000,
    "dtype": torch.float32,
    "cat_coords": 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 Tensor, ndim 3, dtype family float, shape (2, 3, 8) — shape (2, 3, 8) 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
preprocess_color Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/improc.py
code served (permissive licence) · get_code("ba50f39a6d884a2e")
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 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
import numpy as np

torch.manual_seed(0)

# Test with numpy array
x_numpy = np.array([[10, 20, 30, 40], [50, 60, 70, 80]], dtype=np.uint8)

ARGS = {
    "x": x_numpy
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — 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 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
preprocess_color_tf Ran
ran, violated the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/improc.py
code served (permissive licence) · get_code("7a83555254767abf")
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 EagerTensor, ndim 3, dtype family float, shape (2, 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 tensorflow as tf
import numpy as np

def preprocess_color_tf(x):
    import tensorflow as tf
    return tf.cast(x, tf.float32) * 1.0 / 255 - 0.5

# Create a small test input tensor
# The function expects an array-like input that can be cast to float32
# Using uint8 values (0-255 range) as typical for color images
x_input = tf.constant(np.array([
    [[255, 128, 64, 32],
     [16, 200, 100, 50]],
    [[10, 20, 30, 40],
     [100, 150, 200, 250]]
], dtype=np.uint8))

ARGS = {
    "x": x_input
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 3, dtype family float, shape (2, 2, 4) — shape (2, 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
sub2ind Ran
ran, honoured the contract we drafted
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/basic.py
code served (permissive licence) · get_code("49458c1bab9add55")
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 int, ndim 0, dtype family int, shape () 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 numpy as np

torch.manual_seed(0)

# Analyze the function: sub2ind(height, width, y, x)
# This function converts 2D coordinates (y, x) to 1D linear indices
# given image dimensions (height, width)
# All parameters are used in arithmetic operations: y * width + x

height = 5
width = 4
y = 2
x = 1

ARGS = {
    "height": height,
    "width": width,
    "y": y,
    "x": x,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () 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
bilinear_sampler Not yet run
unverified (harvested, no recorded run)
dsl-lab/fvmd-frechet-video-motion-distance/fvmd/nets/pips2.py
code served (permissive licence) · get_code("1ac8cce613d98769")
How it ran

no invocation recorded.

calc_hist Not yet run
unverified (harvested, no recorded run)
ljh0v0/fmd-frechet-motion-distance/fvmd/extract_motion_features.py
code served (permissive licence) · get_code("cf1497316cd1cf12")
How it ran

no invocation recorded.

pca_embed Not yet run
unverified (harvested, no recorded run)
ljh0v0/fmd-frechet-motion-distance/fvmd/utils/improc.py
code served (permissive licence) · get_code("0794073727c4b054")
How it ran

no invocation recorded.

Also present in other papers' code

Of 23 samples lifted from this paper's linked repositories, 23 carry a function identity and 0 do not. Of the 23 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 1 later) and 19 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
bilinear_sampler 18 other papers
13 dated earlier, 5 later · 7 only as an assembly brick
2003.12039 (2020-03)
RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
2108.00622 (2021-08)
Recurrent Mask Refinement for Few-Shot Medical Image Segmentation
2108.10552 (2021-08)
E-RAFT: Dense Optical Flow from Event Cameras
and 15 more not named here
matmul2 2 other papers
2 dated earlier
2303.09551 (2023-03)
SurroundOcc: Multi-Camera 3D Occupancy Prediction for Autonomous Driving
2303.10076 (2023-03)
A Simple Framework for 3D Occupancy Estimation in Autonomous Driving
coords_grid 1 other paper
1 dated earlier
2307.15055 (2023-07)
PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Tracking
load 1 other paper
1 dated earlier
2307.15055 (2023-07)
PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Tracking

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

Significant advancements have been made in video generative models recently. Unlike image generation, video generation presents greater challenges, requiring not only generating high-quality frames but also ensuring temporal consistency across these frames. Despite the impressive progress, research on metrics for evaluating the quality of generated videos, especially concerning temporal and motion consistency, remains underexplored. To bridge this research gap, we propose Fréchet Video Motion Distance (FVMD) metric, which focuses on evaluating motion consistency in video generation. Specifically, we design explicit motion features based on key point tracking, and then measure the similarity between these features via the Fréchet distance. We conduct sensitivity analysis by injecting noise into real videos to verify the effectiveness of FVMD. Further, we carry out a large-scale human study, demonstrating that our metric effectively detects temporal noise and aligns better with human perceptions of generated video quality than existing metrics. Additionally, our motion features can consistently improve the performance of Video Quality Assessment (VQA) models, indicating that our approach is also applicable to unary video quality evaluation. Code is available at https://github.com/ljh0v0/FMD-frechet-motion-distance.

For agents

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

get_harvested_code_for_paper("2407.16124")
get_code_for_paper("2407.16124")
have("2407.16124")

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

curl https://syntology.ai/api/ran/2407.16124.json

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

[![Syntology run record](https://syntology.ai/api/ran/2407.16124.svg)](https://syntology.ai/paper/2407.16124)

Connect an agent — have() is free.