SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1812.11941 · 2018

High Quality Monocular Depth Estimation via Transfer Learning

arXiv · PDF · Open in the Atlas

Code that ran

19 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; 3 honoured the contract we drafted, 0 violated it and 13 have no contract verdict; 3 failures of our instrument; 4 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/1812.11941.json · badge: /api/ran/1812.11941.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
dsshim0125/grmc community (classifier label)
classifier basis not recorded on the link
9 ran · 1 no recorded run
lemon-liley/isabella community (classifier label)
classifier basis not recorded on the link
7 ran · 2 no recorded run
rramjee/DenseDepth reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
KarthikGangadhar/depth-estimation extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
DepthNorm Ran
ran, honoured the contract we drafted
rramjee/DenseDepth/Tensorflow/evaluate.py
pointer only (licence: GPL-3.0) · get_code("1852f926ecf3ccc4")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, rramjee/DenseDepth, is not in the permissive set Syntology serves inline. 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.

DepthNorm Ran
ran on a synthesized input
dsshim0125/grmc/utils.py
code served (permissive licence) · get_code("d57ca4e488622dcb")
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
import numpy as np

torch.manual_seed(0)

# The function DepthNorm takes a depth tensor and divides maxDepth by it
# depth should be a tensor or numeric value with depth values
# maxDepth is a scalar float (default 1000.0)

depth = torch.tensor([[100.0, 200.0, 250.0, 500.0],
                      [150.0, 300.0, 400.0, 600.0]], dtype=torch.float32)

maxDepth = 1000.0

ARGS = {
    'depth': depth,
    'maxDepth': maxDepth
}
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 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
augment Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/inference.py
code served (permissive licence) · get_code("6c798decc8af5b44")
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
import numpy as np

torch.manual_seed(0)

# x_img: a torch tensor that will be converted to numpy
# The function uses x_img.numpy(), then accesses x_img.shape[-1]
# Shape should be (batch, height, width, channels) or similar
# Let's use shape (2, 3, 4, 5) - batch=2, height=3, width=4, channels=5
x_img = torch.randn(2, 3, 4, 5)

# flip: boolean parameter
flip = True

# rotate: iterable of shift proportions (floats between -1 and 1)
rotate = [0.1, -0.25]

ARGS = {
    'x_img': x_img,
    'flip': flip,
    'rotate': rotate
}
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
augment_undo Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/inference.py
code served (permissive licence) · get_code("d7030b9fc8555db4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 5, dtype family float, shape (2, 4, 3, 5, 2) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create augmented images tensor
# The function expects x_imgs_augmented to have a .cpu().numpy() method
# so it should be a torch tensor
# Shape: (batch_size * num_aug_types, height, width, channels)
# We'll use 8 total images (4 images × 2 augmentation types)
batch_size = 4
height = 3
width = 5
channels = 2
num_aug_types = 2

x_imgs_augmented = torch.randn(batch_size * num_aug_types, height, width, channels)

# Define augmentation types that the function will undo
# The function divides the batch by len(aug_type), so we need aug_type to have length 2
aug_type = ['flip', 'rotate 2']

ARGS = {
    'x_imgs_augmented': x_imgs_augmented,
    'aug_type': aug_type
}
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 5, dtype family float, shape (2, 4, 3, 5, 2) — shape (2, 4, 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
create_window Ran
ran, honoured the contract we drafted
dsshim0125/grmc/loss.py
pointer only (licence: MIT); the stored copy get_code serves, from Chhro123/fast-foreground-aware-anomaly-synthesis, is pointer only (licence: NONE) · get_code("443341084ce76798")
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 Chhro123/fast-foreground-aware-anomaly-synthesis, 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.

display_single_image Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
KarthikGangadhar/depth-estimation/video_display.py
code served (permissive licence) · get_code("d84535097384ce79")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned ndarray, ndim 3, dtype family float, shape (5, 12, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import matplotlib.pyplot as plt

# Create a small output array (height=5, width=6, channels=1 for colormap)
output = np.random.rand(5, 6, 1) * 10 + 5  # values between 5 and 15

# Create a small input array (height=5, width=6, channels=3 for RGB)
inputs = np.random.rand(5, 6, 3)

ARGS = {
    'output': output,
    'inputs': inputs,
    'is_colormap': True
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
matplotlib, 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 float, shape (5, 12, 3) — shape (5, 12, 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
extract_zip Ran
ran on a synthesized input
dsshim0125/grmc/evaluate_pytorch.py
code served (permissive licence) · get_code("58071b7eba94ccbd")
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 dict. 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 io
import zipfile
from zipfile import ZipFile

# Create an in-memory zip file
zip_buffer = io.BytesIO()
with ZipFile(zip_buffer, 'w') as zf:
    zf.writestr('file1.txt', b'content of file 1')
    zf.writestr('file2.txt', b'content of file 2')
    zf.writestr('subdir/file3.txt', b'content of file 3')

zip_buffer.seek(0)

ARGS = {
    'input_zip': zip_buffer
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
find_N_peaks Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/inference.py
code served (permissive licence) · get_code("2c99a936790c4965")
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 2, 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

# Create a simple signal with some peaks
# Using a synthetic signal that's periodic so wrap mode makes sense
signal = np.array([0.1, 0.3, 0.15, 0.5, 0.2, 0.8, 0.25, 0.6, 0.12, 0.4], dtype=np.float32)

ARGS = {
    'signal': signal,
    'r': 3,
    'min_v': 0.1,
    'N': 2
}
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 tuple, ndim 2, dtype family float — 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
gaussian Ran
ran, honoured the contract we drafted
dsshim0125/grmc/loss.py
code served (permissive licence) · get_code("c56b7ef16f309a45")
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 1, dtype family float, shape (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
from math import exp

torch.manual_seed(0)

ARGS = {
    "window_size": 5,
    "sigma": 1.0,
}
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 1, dtype family float, shape (5,) — shape (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
getNoTransform Ran
ran on a synthesized input
dsshim0125/grmc/data.py
code served (permissive licence) · get_code("8d2d575309864698")
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 Compose. 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
from PIL import Image

torch.manual_seed(0)

ARGS = {
    'is_test': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Compose — a Compose output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Compose 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
loadZipToMem Ran
ran on a synthesized input
dsshim0125/grmc/data.py
code served (permissive licence) · get_code("744ce2b9aed9a499")
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 io
import zipfile
import csv

# Create a minimal zip file with the required CSV structure in memory
zip_buffer = io.BytesIO()
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zf:
    # Create training data CSV
    train_csv = "id,feature1,feature2\n"
    for i in range(100):
        train_csv += f"{i},value_{i}_a,value_{i}_b\n"
    zf.writestr('data/nyu2_train.csv', train_csv)
    
    # Create test data CSV
    test_csv = "id,feature1,feature2\n"
    for i in range(20):
        test_csv += f"{i},test_{i}_a,test_{i}_b\n"
    zf.writestr('data/nyu2_test.csv', test_csv)

zip_buffer.seek(0)

# Write the zip file to a temporary location (in-memory approach)
import tempfile
import os

temp_dir = tempfile.gettempdir()
zip_file_path = os.path.join(temp_dir, 'test_dataset.zip')

with open(zip_file_path, 'wb') as f:
    f.write(zip_buffer.getvalue())

ARGS = {
    'zip_file': zip_file_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
sklearn — 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
loadZipToMem_test Ran
ran on a synthesized input
dsshim0125/grmc/data.py
code served (permissive licence) · get_code("95089e2ebb03ecf5")
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 tempfile
import os
from zipfile import ZipFile

# Create a temporary zip file with the required structure
temp_dir = tempfile.mkdtemp()
zip_path = os.path.join(temp_dir, 'test_dataset.zip')

# Create test CSV data
csv_content = """image1.jpg,label1
image2.jpg,label2
image3.jpg,label3
image4.jpg,label4
image5.jpg,label5"""

# Create the zip file with the expected structure
with ZipFile(zip_path, 'w') as zf:
    zf.writestr('data/nyu2_test.csv', csv_content)

ARGS = {
    'zip_file': zip_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
sklearn — 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
lr_pad Ran
ran; our contract draft was wrong, not the code
lemon-liley/isabella/HorizonNet/model.py
code served (permissive licence) · get_code("fb884f33b5fda8ea")
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, 4, 7) 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 tensor with shape (batch=2, channels=3, height=4, width=5)
# The function pads along dim=3 (width dimension)
x = torch.randn(2, 3, 4, 5)

# padding parameter
padding = 1

ARGS = {
    "x": x,
    "padding": padding,
}
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, 4, 7) — shape (2, 3, 4, 7) 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
predict Ran
ran; our contract draft was wrong, not the code
rramjee/DenseDepth/Tensorflow/evaluate.py
pointer only (licence: GPL-3.0) · get_code("6ce438ff124a45a8")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, rramjee/DenseDepth, is not in the permissive set Syntology serves inline. 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.

predict Ran
ran on a synthesized input
dsshim0125/grmc/utils.py
code served (permissive licence) · get_code("9637ab8ac25f366c")
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 4, dtype family float, shape (1, 4, 3, 1) 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 numpy as np
import torch
import torch.nn as nn

torch.manual_seed(0)

# Build a minimal model that accepts (batch, channels, height, width) 
# and outputs (batch, 1, height, width) as expected by the predict function
class MinimalDepthModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv = nn.Conv2d(3, 1, kernel_size=3, padding=1)
    
    def forward(self, x):
        return self.conv(x)

# Create model instance
model = MinimalDepthModel()

# Create a small test image: (4, 3) grayscale which will be stacked to (4, 3, 3)
images = np.random.rand(4, 3).astype(np.float32)

ARGS = {
    "model": model,
    "images": images,
    "minDepth": 10,
    "maxDepth": 1000,
    "batch_size": 2
}
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 4, dtype family float, shape (1, 4, 3, 1) — shape (1, 4, 3, 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: none by the harness; any seed is in the setup block
sort_xy_filter_unique Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/dataset.py
code served (permissive licence) · get_code("87dffac4821848bf")
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 2, 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

# Analyze the function:
# - xs: array-like, used with np.array(), np.argsort(), np.unique(), np.diff()
# - ys: array-like, used with np.array(), ys.max(), indexing
# - y_small_first: bool, default True, used in arithmetic to produce 1 or -1

# Create valid test data
xs = np.array([3.0, 1.0, 2.0])  # unsorted values
ys = np.array([10.0, 5.0, 8.0])  # corresponding y values

ARGS = {
    "xs": xs,
    "ys": ys,
    "y_small_first": True,
}
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, ndim 2, dtype family float — 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
ssim Ran
ran on a synthesized input
dsshim0125/grmc/loss.py
code served (permissive licence) · get_code("983373e062ffdb0f")
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 small test images: batch_size=2, channels=3, height=4, width=4
img1 = torch.randn(2, 3, 4, 4)
img2 = torch.randn(2, 3, 4, 4)

# Value range for normalization
val_range = 1.0

# Window size for SSIM computation
window_size = 11

# Window can be None (will be created in function) or pre-created
window = None

# Whether to average the result
size_average = True

# Whether to return full output (ssim + cs) or just ssim
full = False

ARGS = {
    'img1': img1,
    'img2': img2,
    'val_range': val_range,
    'window_size': window_size,
    'window': window,
    'size_average': size_average,
    'full': full
}
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
tri2halfspace Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/eval_cuboid.py
code served (permissive licence) · get_code("7525fc8039cc75b5")
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 1, 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

# The function tri2halfspace takes three 3D points (pa, pb, p)
# and computes a halfspace representation based on their cross product.
# All parameters should be 1D numpy arrays of length 3 (3D points).

pa = np.array([1.0, 2.0, 3.0])
pb = np.array([4.0, 5.0, 6.0])
p = np.array([0.0, 0.0, 0.0])

ARGS = {
    "pa": pa,
    "pb": pb,
    "p": p
}
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 list, ndim 1, dtype family float — 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
xyzlst2halfspaces Ran
ran on a synthesized input
lemon-liley/isabella/HorizonNet/eval_cuboid.py
code served (permissive licence) · get_code("373920e32dc5fdb1")
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 (24, 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

# Create small test data for xyzlst2halfspaces
# The function expects two 2D arrays of 3D points
# xyz_floor: N x 3 array of floor points
# xyz_ceil: N x 3 array of ceiling points

# Create a simple polygon with 4 vertices (N=4)
xyz_floor = np.array([
    [1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0],
    [-1.0, 0.0, 0.0],
    [0.0, -1.0, 0.0]
], dtype=np.float64)

xyz_ceil = np.array([
    [1.0, 0.0, 2.0],
    [0.0, 1.0, 2.0],
    [-1.0, 0.0, 2.0],
    [0.0, -1.0, 2.0]
], dtype=np.float64)

ARGS = {
    "xyz_floor": xyz_floor,
    "xyz_ceil": xyz_ceil
}
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 2, dtype family float, shape (24, 4) — shape (24, 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
choose_decoder Not yet run
unverified (harvested, no recorded run)
dsshim0125/grmc/model/fcrn.py
code served (permissive licence) · get_code("b3d7054d301b12b6")
How it ran

no invocation recorded.

depth_loss_function Not yet run
unverified (harvested, no recorded run)
lemon-liley/isabella/loss.py
code served (permissive licence) · get_code("dddbd5a319a669ff")
How it ran

no invocation recorded.

make_image Not yet run
unverified (harvested, no recorded run)
lemon-liley/isabella/callbacks.py
code served (permissive licence) · get_code("a5c4a0adfc21338d")
How it ran

no invocation recorded.

scale_up Not yet run
unverified (harvested, no recorded run)
rramjee/DenseDepth/Tensorflow/evaluate.py
pointer only (licence: GPL-3.0) · get_code("229418e47c5aa9bc")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, rramjee/DenseDepth, is not in the permissive set Syntology serves inline. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 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, 3 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 3 later) and 20 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
gaussian 160 other papers
160 later · 7 only as an assembly brick
1904.01538 (2019-04)
Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset
1907.11628 (2019-07)
Unsupervised Learning for Optical Flow Estimation Using Pyramid Convolution LSTM
2002.11616 (2020-02)
Zooming Slow-Mo: Fast and Accurate One-Stage Space-Time Video Super-Resolution
and 157 more not named here
create_window 19 other papers
19 later
Zavrtanik_DRAEM_-_A_Discriminatively_Trained_Reconstruction_Embedding_for_Surface_Anomaly_ICCV_2021_paper (2021 (year only))
DRAEM -A discriminatively trained reconstruction embedding for surface anomaly detection
2108.07610 (2021-08)
DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detection
2111.15666 (2021-11)
HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing
and 16 more not named here
lr_pad 2 other papers
2 later
2104.00568 (2021-04)
LED2-Net: Monocular 360 Layout Estimation via Differentiable Depth Rendering
2203.01824 (2022-03)
LGT-Net: Indoor Panoramic Room Layout Estimation with Geometry-Aware Transformer Network

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

More are linked than shown here.

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

Abstract

Accurate depth estimation from images is a fundamental task in many applications including scene understanding and reconstruction. Existing solutions for depth estimation often produce blurry approximations of low resolution. This paper presents a convolutional neural network for computing a high-resolution depth map given a single RGB image with the help of transfer learning. Following a standard encoder-decoder architecture, we leverage features extracted using high performing pre-trained networks when initializing our encoder along with augmentation and training strategies that lead to more accurate results. We show how, even for a very simple decoder, our method is able to achieve detailed high-resolution depth maps. Our network, with fewer parameters and training iterations, outperforms state-of-the-art on two datasets and also produces qualitatively better results that capture object boundaries more faithfully. Code and corresponding pre-trained weights are made publicly available.

For agents

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

get_harvested_code_for_paper("1812.11941")
get_code_for_paper("1812.11941")
have("1812.11941")

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

curl https://syntology.ai/api/ran/1812.11941.json

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

[![Syntology run record](https://syntology.ai/api/ran/1812.11941.svg)](https://syntology.ai/paper/1812.11941)

Connect an agent — have() is free.