SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2008.07130 · 2020

Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation

arXiv · PDF · Open in the Atlas

Code that ran

5 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 failures of our instrument; 10 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/2008.07130.json · badge: /api/ran/2008.07130.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
FilippoAleotti/Reversing canonical (classifier label)
classifier basis: census
5 ran · 10 no recorded run
FunctionStatusWhere it lives
SSIM Ran
ran on a synthesized input
FilippoAleotti/Reversing/mono/losses.py
code served (permissive licence) · get_code("6d143dc92e7f5729")
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 4, dtype family float, shape (2, 3, 3, 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 tensorflow as tf
import numpy as np

# Create small test tensors with shape BxHxWxC
# B=2 (batch), H=5, W=5, C=3 (channels)
# Using float32 as TensorFlow expects
batch_size = 2
height = 5
width = 5
channels = 3

# Create random tensors with distinct dimensions
np.random.seed(0)
x_data = np.random.randn(batch_size, height, width, channels).astype(np.float32)
y_data = np.random.randn(batch_size, height, width, channels).astype(np.float32)

x = tf.constant(x_data)
y = tf.constant(y_data)

ARGS = {
    'x': x,
    'y': y,
}
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 4, dtype family float, shape (2, 3, 3, 3) — shape (2, 3, 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
bilinear_sampler Ran
ran on a synthesized input
FilippoAleotti/Reversing/mono/bilinear_sampler.py
code served (permissive licence) · get_code("c446755f0a857957")
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 4, dtype family float, shape (2, 4, 3, 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 tensorflow as tf
import numpy as np

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

# Create a small batch of images
# Shape: [batch_size, height, width, channels]
batch_size = 2
height = 4
width = 3
channels = 5

imgs = tf.constant(
    np.random.randn(batch_size, height, width, channels).astype(np.float32)
)

# Create coordinate tensors
# Shape: [batch_size, height, width, 2] where last dim is [x, y] coordinates
coords = tf.constant(
    np.random.uniform(0, width-1, size=(batch_size, height, width, 2)).astype(np.float32)
)

ARGS = {
    'imgs': imgs,
    'coords': coords,
}
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 4, dtype family float, shape (2, 4, 3, 5) — shape (2, 4, 3, 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
build_coords Ran
ran on a synthesized input
FilippoAleotti/Reversing/mono/bilinear_sampler.py
code served (permissive licence) · get_code("5c8abcee36f23e1e")
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 4, dtype family float, shape (2, 3, 4, 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 tensorflow as tf
import numpy as np

# Create a small input tensor with shape (batch=2, height=3, width=4, channels=1)
# The function expects coords to have get_shape() method that returns dimensions
coords_input = tf.constant(np.random.randn(2, 3, 4, 1), dtype=tf.float32)

ARGS = {
    'coords': coords_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 4, dtype family float, shape (2, 3, 4, 2) — shape (2, 3, 4, 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
generate_image_left Ran
ran on a synthesized input
FilippoAleotti/Reversing/mono/bilinear_sampler.py
code served (permissive licence) · get_code("cc9d1daeffdd0820")
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 4, dtype family float, shape (2, 4, 3, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf
import numpy as np

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

# Create tiny test tensors
batch_size = 2
height = 4
width = 3
channels = 1

# Create a small image tensor (batch, height, width, channels)
img = tf.constant(np.random.rand(batch_size, height, width, channels).astype(np.float32))

# Create a small disparity tensor (batch, height, width, 1)
# Disparity values should be small offsets (typically fractional pixel displacements)
disp = tf.constant(np.random.rand(batch_size, height, width, 1).astype(np.float32) * 0.5)

ARGS = {
    'img': img,
    'disp': disp,
}
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 4, dtype family float, shape (2, 4, 3, 1) — shape (2, 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
gradient_x Ran
ran on a synthesized input
FilippoAleotti/Reversing/mono/losses.py
code served (permissive licence) · get_code("9460bd1f0f3ca10f")
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 4, dtype family float, shape (2, 4, 3, 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 tensorflow as tf
import numpy as np

# Create a small batch of images
# Shape: (batch_size, height, width, channels)
# Using distinct sizes: batch=2, height=4, width=3, channels=5
batch_size = 2
height = 4
width = 3
channels = 5

# Create a random image tensor with shape (batch, height, width, channels)
img = tf.constant(np.random.randn(batch_size, height, width, channels).astype(np.float32))

ARGS = {
    "img": img
}
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 4, dtype family float, shape (2, 4, 3, 5) — shape (2, 4, 3, 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
berhu_loss Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/losses.py
code served (permissive licence) · get_code("741e7cb6b7e04a6c")
How it ran

no invocation recorded.

colorize Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/tools.py
code served (permissive licence) · get_code("76dcde6460136af6")
How it ran

no invocation recorded.

conv2d Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/ops.py
code served (permissive licence) · get_code("20792d996729d17a")
How it ran

no invocation recorded.

conv2d_transpose Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/ops.py
code served (permissive licence) · get_code("bf9685b92676e2dc")
How it ran

no invocation recorded.

correlation_map Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/ops.py
code served (permissive licence) · get_code("1c98a5d5a6ae33ad")
How it ran

no invocation recorded.

count_text_lines Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/tools.py
code served (permissive licence) · get_code("e74ad6c7bd167a43")
How it ran

no invocation recorded.

disparity_estimation Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/monoResMatch.py
code served (permissive licence) · get_code("8053656f836ca9e5")
How it ran

no invocation recorded.

disparity_refinement Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/monoResMatch.py
code served (permissive licence) · get_code("aa6d420099a97869")
How it ran

no invocation recorded.

stem_block Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/monoResMatch.py
code served (permissive licence) · get_code("fc9baaf5c13a2f4a")
How it ran

no invocation recorded.

string_length_tf Not yet run
unverified (harvested, no recorded run)
FilippoAleotti/Reversing/mono/dataloader.py
code served (permissive licence) · get_code("f4a175127ff1362b")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 15 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

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

Repositories linked to this paper

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

Abstract

In many fields, self-supervised learning solutions are rapidly evolving and filling the gap with supervised approaches. This fact occurs for depth estimation based on either monocular or stereo, with the latter often providing a valid source of self-supervision for the former. In contrast, to soften typical stereo artefacts, we propose a novel self-supervised paradigm reversing the link between the two. Purposely, in order to train deep stereo networks, we distill knowledge through a monocular completion network. This architecture exploits single-image clues and few sparse points, sourced by traditional stereo algorithms, to estimate dense yet accurate disparity maps by means of a consensus mechanism over multiple estimations. We thoroughly evaluate with popular stereo datasets the impact of different supervisory signals showing how stereo networks trained with our paradigm outperform existing self-supervised frameworks. Finally, our proposal achieves notable generalization capabilities dealing with domain shift issues. Code available at https://github.com/FilippoAleotti/Reversing

For agents

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

get_harvested_code_for_paper("2008.07130")
get_code_for_paper("2008.07130")
have("2008.07130")

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

curl https://syntology.ai/api/ran/2008.07130.json

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

[![Syntology run record](https://syntology.ai/api/ran/2008.07130.svg)](https://syntology.ai/paper/2008.07130)

Connect an agent — have() is free.