SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1909.11895 · 2019

Joint-task Self-supervised Learning for Temporal Correspondence

arXiv · PDF · Open in the Atlas

Code that ran

18 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; 1 honoured the contract we drafted, 0 violated it and 14 have no contract verdict; 3 failures of our instrument; 5 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1909.11895.json · badge: /api/ran/1909.11895.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
Liusifei/UVC canonical (classifier label)
classifier basis: census
18 ran · 5 no recorded run
FunctionStatusWhere it lives
adjust_learning_rate Ran
ran on a synthesized input
Liusifei/UVC/track_match_v1.py
code served (permissive licence) · get_code("efe9d4fd3e733289")
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 float, ndim 0, dtype family float, 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 torch.optim as optim

torch.manual_seed(0)

# Create a simple model
class SimpleModel(nn.Module):
    def __init__(self):
        super(SimpleModel, self).__init__()
        self.linear = nn.Linear(4, 2)
    
    def forward(self, x):
        return self.linear(x)

model = SimpleModel()
optimizer = optim.SGD(model.parameters(), lr=0.1)

# Create args object with required attributes
class Args:
    def __init__(self):
        self.lr = 0.1
        self.lr_mode = 'step'
        self.step = 30
        self.nepoch = 100

args = Args()
epoch = 15

ARGS = {
    'args': args,
    'optimizer': optimizer,
    'epoch': epoch
}
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 float, ndim 0, dtype family float, 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
conv1x1 Ran
ran; our contract draft was wrong, not the code
Liusifei/UVC/libs/resnet.py
code served (permissive licence) · get_code("d9def42110729a85")
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 Conv2d. 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)

# The function conv1x1 takes in_planes, out_planes, and stride as parameters
# It returns an nn.Conv2d layer configured for 1x1 convolution
# We need to provide valid integer arguments for these parameters

ARGS = {
    "in_planes": 3,
    "out_planes": 64,
    "stride": 1
}
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 Conv2d — a Conv2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Conv2d 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
conv3x3 Ran
ran; our contract draft was wrong, not the code
Liusifei/UVC/libs/resnet.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.

coords2bbox Ran
ran on a synthesized input
Liusifei/UVC/model.py
code served (permissive licence) · get_code("c3e082f966f8d659")
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, 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)

# coords: shape (batch_size, num_points, 2) - coordinates of pixels
# patch_size: integer - patch size (not directly used in function body)
# h_tar: integer - target image height
# w_tar: integer - target image width

batch_size = 2
num_points = 4
coords = torch.randn(batch_size, num_points, 2) * 100  # Random coordinates
coords = torch.abs(coords)  # Make positive

patch_size = 16
h_tar = 480
w_tar = 640

ARGS = {
    "coords": coords,
    "patch_size": patch_size,
    "h_tar": h_tar,
    "w_tar": w_tar,
}
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, 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
create_flat_grid Ran
ran on a synthesized input
Liusifei/UVC/model.py
code served (permissive licence) · get_code("c032af918144806e")
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, 15, 2) 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)

# F_size is a tuple (b, c, h, w) representing batch size, channels, height, width
F_size = (2, 4, 3, 5)

ARGS = {
    "F_size": F_size,
    "GPU": 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, 15, 2) — shape (2, 15, 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
create_grid Ran
ran on a synthesized input
Liusifei/UVC/libs/concentration_loss.py
code served (permissive licence) · get_code("58402808d405db23")
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, 5, 2) 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
import torch.nn as nn

torch.manual_seed(0)

# F_size is a tuple (b, c, h, w) representing batch, channels, height, width
F_size = (2, 3, 4, 5)

# GPU parameter - set to False since we're CPU-only
GPU = False

ARGS = {
    "F_size": F_size,
    "GPU": GPU,
}
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, 5, 2) — shape (2, 4, 5, 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
create_grid Ran
ran on a synthesized input
Liusifei/UVC/libs/track_utils.py
code served (permissive licence) · get_code("2bfe1a16c66739f6")
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, 3, 5, 2) 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)

# F_size is a tuple of (batch, channels, height, width)
# The function uses it to create an affine grid, so it needs to be a valid 4-tuple
F_size = (2, 4, 3, 5)

# GPU parameter is a boolean - set to False since we're CPU-only
GPU = False

ARGS = {
    "F_size": F_size,
    "GPU": GPU,
}
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, 3, 5, 2) — shape (2, 3, 5, 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
draw_bbox Ran
ran on a synthesized input
Liusifei/UVC/libs/train_utils.py
code served (permissive licence) · get_code("a5d852d21a97fde4")
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 (100, 150, 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 cv2
import numpy as np

# Create a simple test image: h * w * 3 numpy array
# Using 100x150x3 to have distinct height and width
img = np.ones((100, 150, 3), dtype=np.uint8) * 200

# Create a bounding box: [left, right, top, bottom]
# Using values that fit within the image bounds
bbox = np.array([10, 120, 20, 80], dtype=np.float32)

ARGS = {
    "img": img,
    "bbox": bbox
}
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 (100, 150, 3) — shape (100, 150, 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
fast_hist Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Liusifei/UVC/eval_vip.py
code served (permissive licence) · get_code("d86fb168246b16a4")
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 ndarray, ndim 2, dtype family int, shape (4, 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 numpy as np

# Analyze the function:
# - a: array-like, used with >= 0, < n comparisons, and .astype(int)
# - b: array-like, same shape as a, used in indexing
# - n: int, used as upper bound and for bincount minlength

# Create small numpy arrays for a and b
a = np.array([0, 1, 2, 3, 1, 2], dtype=np.int32)
b = np.array([0, 1, 2, 0, 1, 2], dtype=np.int32)
n = 4

ARGS = {
    "a": a,
    "b": b,
    "n": n
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 2, dtype family int, shape (4, 4) — shape (4, 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
featureL2Norm Ran
ran, honoured the contract we drafted
Liusifei/UVC/libs/net_utils.py
code served (permissive licence) · get_code("cf9656ffec66503c")
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 a small feature tensor with shape (batch_size=2, feature_dim=4)
# The function expects a 2D tensor where dimension 1 is summed over
feature = torch.randn(2, 4)

ARGS = {
    "feature": feature
}
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
framepair_loader Ran
ran on a synthesized input
Liusifei/UVC/libs/loader.py
code served (permissive licence) · get_code("e15375d29905159e")
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, ndim 4, 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 numpy as np
import random
import cv2
import tempfile
import os

# Create a temporary video file for testing
def create_dummy_video(filepath, num_frames=100, width=128, height=128):
    """Create a dummy video file with specified parameters."""
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
    out = cv2.VideoWriter(filepath, fourcc, 30.0, (width, height))
    
    for i in range(num_frames):
        # Create a simple frame with changing colors
        frame = np.zeros((height, width, 3), dtype=np.uint8)
        frame[:, :] = [i % 256, (i * 2) % 256, (i * 3) % 256]
        out.write(frame)
    
    out.release()

# Create a temporary directory and video file
temp_dir = tempfile.mkdtemp()
video_path = os.path.join(temp_dir, "test_video.mp4")
create_dummy_video(video_path, num_frames=100, width=128, height=128)

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

ARGS = {
    "video_path": video_path,
    "frame_start": 10,
    "frame_end": 90
}
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 list, ndim 4, dtype family int — a list 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
resnet18 Ran
ran on a synthesized input
Liusifei/UVC/libs/resnet.py
code served (permissive licence) · get_code("1ccbb405b8feeb4f")
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 ResNet. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'pretrained': 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 ResNet — a ResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet 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
switch_L1_thr Ran
ran on a synthesized input
Liusifei/UVC/libs/loss.py
code served (permissive licence) · get_code("13a19c9c06357b8b")
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 0, dtype family float, 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

torch.manual_seed(0)

# Create tiny tensors with shape (batch=2, channels=3, height=4, width=4)
# pred2, frame2_var, pred1, frame1_var are all 4D tensors used in weightsingle_L1
pred2 = torch.randn(2, 3, 4, 4)
frame2_var = torch.randn(2, 3, 4, 4)
pred1 = torch.randn(2, 3, 4, 4)
frame1_var = torch.randn(2, 3, 4, 4)

# Scalar weights for a and b channels
w_a = 0.5
w_b = 0.3

# Threshold value for filtering losses
thr = 0.1

ARGS = {
    "pred2": pred2,
    "frame2_var": frame2_var,
    "pred1": pred1,
    "frame1_var": frame1_var,
    "w_a": w_a,
    "w_b": w_b,
    "thr": thr,
}
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 0, dtype family float, 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
to_one_hot Ran
ran on a synthesized input
Liusifei/UVC/libs/track_utils.py
code served (permissive licence) · get_code("64adabd58edb62ca")
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 (1, 9, 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)

# The function expects a tensor with shape (C, H, W) where it unpacks as _, h, w = y_tensor.size()
# So we need a 3D tensor. The values should be integers in range [0, n_dims-1]
# Let's create a small tensor with shape (1, 4, 3) - 1 channel, 4 height, 3 width
y_tensor = torch.randint(0, 9, (1, 4, 3))

ARGS = {
    "y_tensor": y_tensor,
    "n_dims": 9
}
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 (1, 9, 4, 3) — shape (1, 9, 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
video_frame_counter Ran
ran on a synthesized input
Liusifei/UVC/libs/loader.py
code served (permissive licence) · get_code("bfd0529a5407cf64")
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 float, ndim 0, dtype family float, 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 cv2
import numpy as np
import tempfile
import os

# Create a temporary video file for testing
def create_test_video(path, num_frames=10):
    """Create a simple test video file"""
    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
    out = cv2.VideoWriter(path, fourcc, 30.0, (640, 480))
    
    for i in range(num_frames):
        # Create a simple frame (640x480 BGR image)
        frame = np.zeros((480, 640, 3), dtype=np.uint8)
        # Add some variation to make frames different
        frame[:, :, 0] = (i * 25) % 256  # Blue channel
        frame[:, :, 1] = (i * 50) % 256  # Green channel
        frame[:, :, 2] = (i * 100) % 256  # Red channel
        out.write(frame)
    
    out.release()

# Create temporary video file
temp_dir = tempfile.gettempdir()
video_path = os.path.join(temp_dir, "test_video_counter.mp4")
create_test_video(video_path, num_frames=10)

ARGS = {
    "video_path": video_path
}
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 float, ndim 0, dtype family float, 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
video_loader Ran
ran on a synthesized input
Liusifei/UVC/libs/loader.py
code served (permissive licence) · get_code("01111e0ea262be80")
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, ndim 4, 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 cv2
import numpy as np
import tempfile
import os

# Create a temporary video file for testing
temp_dir = tempfile.gettempdir()
video_path = os.path.join(temp_dir, "test_video.mp4")

# Create a simple test video using OpenCV
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter(video_path, fourcc, 20.0, (640, 480))

# Write 10 frames to the video
for i in range(10):
    frame = np.zeros((480, 640, 3), dtype=np.uint8)
    # Add some variation to each frame so we can distinguish them
    frame[:, :] = [i * 25 % 256, (i * 50) % 256, (i * 75) % 256]
    out.write(frame)

out.release()

# Define arguments for video_loader
ARGS = {
    'video_path': video_path,
    'frame_end': 8,
    'step': 2,
    'frame_start': 1
}
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 list, ndim 4, dtype family int — a list 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
weight_L1 Ran
ran on a synthesized input
Liusifei/UVC/libs/loss.py
code served (permissive licence) · get_code("9d2097efd0543d8a")
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 0, dtype family float, 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

torch.manual_seed(0)

# Create pred and gt tensors with shape (batch=2, channels=3, height=4, width=5)
# The function accesses pred[:, 0, :, :], pred[:, 1, :, :], pred[:, 2, :, :]
# so we need exactly 3 channels
pred = torch.randn(2, 3, 4, 5)
gt = torch.randn(2, 3, 4, 5)

# Create mask with shape (batch=2, 1, height=4, width=5)
# It gets repeated to match pred.size(1) which is 3
mask = torch.ones(2, 1, 4, 5)

# Weight parameters
w_a = 0.5
w_b = 0.3

ARGS = {
    "pred": pred,
    "gt": gt,
    "w_a": w_a,
    "w_b": w_b,
    "mask": mask
}
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 0, dtype family float, 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
weightsingle_L1 Ran
ran on a synthesized input
Liusifei/UVC/libs/loss.py
code served (permissive licence) · get_code("facf9894c353245d")
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

torch.manual_seed(0)

# Create small tensors for pred and gt
# pred and gt should have shape (n, c, h, w) where c=3 (L, a, b channels)
batch_size = 2
channels = 3
height = 4
width = 5

pred = torch.randn(batch_size, channels, height, width)
gt = torch.randn(batch_size, channels, height, width)

# Weight parameters for a and b channels
w_a = 0.5
w_b = 0.3

ARGS = {
    "pred": pred,
    "gt": gt,
    "w_a": w_a,
    "w_b": w_b,
}
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
aff2coord Not yet run
unverified (harvested, no recorded run)
Liusifei/UVC/libs/concentration_loss.py
code served (permissive licence) · get_code("9058715f5ac3cb5b")
How it ran

no invocation recorded.

compute_hist Not yet run
unverified (harvested, no recorded run)
Liusifei/UVC/eval_vip.py
pointer only (licence: MIT); the stored copy get_code serves, from HCPLab-SYSU/ATEN, is pointer only (licence: NOASSERTION) · get_code("384eac2f5e92af52")
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 HCPLab-SYSU/ATEN, 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.

sample_patch Not yet run
unverified (harvested, no recorded run)
Liusifei/UVC/libs/train_utils.py
code served (permissive licence) · get_code("fcd7eaa2c70cfcd8")
How it ran

no invocation recorded.

seg2bbox_v2 Not yet run
unverified (harvested, no recorded run)
Liusifei/UVC/libs/track_utils.py
code served (permissive licence) · get_code("a336a612bbbc899a")
How it ran

no invocation recorded.

transform Not yet run
unverified (harvested, no recorded run)
Liusifei/UVC/model.py
code served (permissive licence) · get_code("98d9c87973a0354e")
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, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 2 the same month, 4 later) and 18 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
154 dated earlier, 8 the same month, 524 later · 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
conv1x1 431 other papers
44 dated earlier, 1 the same month, 385 later, 1 undated · 38 only as an assembly brick
1409.1556 (2014-09)
Very Deep Convolutional Networks for Large-Scale Image Recognition
1506.01186 (2015-06)
Cyclical Learning Rates for Training Neural Networks
1507.05717 (2015-07)
An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition
and 428 more not named here
fast_hist 12 other papers
6 dated earlier, 6 later
1511.00561 (2015-11)
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation
1702.01105 (2017-02)
Joint 2D-3D-Semantic Data for Indoor Scene Understanding
1806.07373 (2018-06)
Few-Shot Segmentation Propagation with Guided Networks
and 9 more not named here
featureL2Norm 7 other papers
2 dated earlier, 5 later · 2 only as an assembly brick
1703.05593 (2017-03)
Convolutional neural network architecture for geometric matching
1810.10510 (2018-10)
Neighbourhood Consensus Networks
2003.12059 (2020-03)
Correspondence Networks with Adaptive Neighbourhood Consensus
and 4 more not named here
compute_hist 1 other paper
1 dated earlier
1808.00661 (2018-08)
Adaptive Temporal Encoding Network for Video Instance-level Human Parsing

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 proposes to learn reliable dense correspondence from videos in a self-supervised manner. Our learning process integrates two highly related tasks: tracking large image regions \emph{and} establishing fine-grained pixel-level associations between consecutive video frames. We exploit the synergy between both tasks through a shared inter-frame affinity matrix, which simultaneously models transitions between video frames at both the region- and pixel-levels. While region-level localization helps reduce ambiguities in fine-grained matching by narrowing down search regions; fine-grained matching provides bottom-up features to facilitate region-level localization. Our method outperforms the state-of-the-art self-supervised methods on a variety of visual correspondence tasks, including video-object and part-segmentation propagation, keypoint tracking, and object tracking. Our self-supervised method even surpasses the fully-supervised affinity feature representation obtained from a ResNet-18 pre-trained on the ImageNet.

For agents

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

get_harvested_code_for_paper("1909.11895")
get_code_for_paper("1909.11895")
have("1909.11895")

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

curl https://syntology.ai/api/ran/1909.11895.json

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

[![Syntology run record](https://syntology.ai/api/ran/1909.11895.svg)](https://syntology.ai/paper/1909.11895)

Connect an agent — have() is free.