SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2008.12977 · 2020

Improved anomaly detection by training an autoencoder with skip connections on images corrupted with Stain-shaped noise

arXiv · PDF · Open in the Atlas

Code that ran

4 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 failures of our instrument; 6 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/2008.12977.json · badge: /api/ran/2008.12977.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
anncollin/AnomalyDetection-Keras canonical (classifier label)
classifier basis: census
4 ran · 6 no recorded run
FunctionStatusWhere it lives
MSE Ran
ran on a synthesized input
anncollin/AnomalyDetection-Keras/Models/Losses.py
code served (permissive licence) · get_code("cd3673bd44747ec7")
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 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 tensorflow as tf
import numpy as np

# Create small tensors with distinct shapes
y_true = tf.constant([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]], dtype=tf.float32)
y_pred = tf.constant([[1.1, 2.1, 2.9, 4.2], [4.9, 6.1, 7.1, 7.9]], dtype=tf.float32)

ARGS = {
    "y_true": y_true,
    "y_pred": y_pred,
}
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 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
PSNRLoss Ran
ran on a synthesized input
anncollin/AnomalyDetection-Keras/Models/Losses.py
code served (permissive licence) · get_code("35960f52558880f0")
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 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 tensorflow as tf
import numpy as np

# Create small image tensors for PSNR calculation
# PSNR expects float values, typically in range [0, 1] or [0, 255]
# We'll use [0, 1] range as specified by max_val=1.0 in the function

batch_size = 2
height = 4
width = 3
channels = 3

# Create y_true: ground truth images
y_true = tf.constant(
    np.random.uniform(0, 1, (batch_size, height, width, channels)).astype(np.float32)
)

# Create y_pred: predicted images (similar but slightly different)
y_pred = tf.constant(
    np.random.uniform(0, 1, (batch_size, height, width, channels)).astype(np.float32)
)

ARGS = {
    "y_true": y_true,
    "y_pred": y_pred,
}
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 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
show_im Ran
ran on a synthesized input
anncollin/AnomalyDetection-Keras/GUI/helper.py
code served (permissive licence) · get_code("328b2554e4914e03")
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 Figure. 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 matplotlib import pyplot as plt

torch.manual_seed(0)

# Create a small grayscale image
# The function expects img to be array-like that can be passed to plt.imshow()
img = np.random.rand(100, 150).astype(np.float32)

# row and col are used for figsize calculation: (col / my_dpi, row / my_dpi)
row = 200
col = 300

ARGS = {
    'img': img,
    'row': row,
    'col': col
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Figure — a Figure output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Figure 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
update_list Ran
ran on a synthesized input
anncollin/AnomalyDetection-Keras/GUI/helper.py
code served (permissive licence) · get_code("85da5760c11b9ea6")
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 torch

torch.manual_seed(0)

# The function filters a list of strings based on query conditions
# - query is a list of query strings
# - each query string can start with '~' (negation) or be a positive match
# - elements are kept if all query conditions are satisfied

list_of_strings = [
    "hello world",
    "python programming",
    "data science",
    "machine learning",
]

query_conditions = [
    "python",
    "~learning",
]

ARGS = {
    "list": list_of_strings,
    "query": query_conditions,
}
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 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
build_model Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/Models/Networks/AE.py
code served (permissive licence) · get_code("08d0ca090165df47")
How it ran

no invocation recorded.

build_model Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/Models/Networks/AESc.py
code served (permissive licence) · get_code("25db04194350d92f")
How it ran

no invocation recorded.

compute_ROC Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/utils/ROC.py
code served (permissive licence) · get_code("49cce9267f9d0497")
How it ran

no invocation recorded.

hjson_parser Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/GUI/helper.py
code served (permissive licence) · get_code("bab1ea2f26dd4acc")
How it ran

no invocation recorded.

l_metric Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/utils/ROC.py
code served (permissive licence) · get_code("78f94b2cd80c9021")
How it ran

no invocation recorded.

psnr Not yet run
unverified (harvested, no recorded run)
anncollin/AnomalyDetection-Keras/utils/ROC.py
code served (permissive licence) · get_code("72e574159ddc79e9")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 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 10 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 industrial vision, the anomaly detection problem can be addressed with an autoencoder trained to map an arbitrary image, i.e. with or without any defect, to a clean image, i.e. without any defect. In this approach, anomaly detection relies conventionally on the reconstruction residual or, alternatively, on the reconstruction uncertainty. To improve the sharpness of the reconstruction, we consider an autoencoder architecture with skip connections. In the common scenario where only clean images are available for training, we propose to corrupt them with a synthetic noise model to prevent the convergence of the network towards the identity mapping, and introduce an original Stain noise model for that purpose. We show that this model favors the reconstruction of clean images from arbitrary real-world images, regardless of the actual defects appearance. In addition to demonstrating the relevance of our approach, our validation provides the first consistent assessment of reconstruction-based methods, by comparing their performance over the MVTec AD dataset, both for pixel- and image-wise anomaly detection.

For agents

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

get_harvested_code_for_paper("2008.12977")
get_code_for_paper("2008.12977")
have("2008.12977")

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

curl https://syntology.ai/api/ran/2008.12977.json

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

[![Syntology run record](https://syntology.ai/api/ran/2008.12977.svg)](https://syntology.ai/paper/2008.12977)

Connect an agent — have() is free.