SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1912.08628 · 2019

Unsupervised Change Detection in Multi-temporal VHR Images Based on Deep Kernel PCA Convolutional Mapping Network

arXiv · PDF · Open in the Atlas

Code that ran

10 of 12 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 10 have no contract verdict; 0 failures of our instrument; 2 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/1912.08628.json · badge: /api/ran/1912.08628.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
I-Hope-Peace/ChangeDetectionRepository community (classifier label)
classifier basis not recorded on the link
10 ran · 2 no recorded run
FunctionStatusWhere it lives
KI Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/cluster_util.py
code served (permissive licence) · get_code("7018126ab22124ba")
How it ran

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

import numpy as np

# Create sample data for the KI function
# The function expects a numpy array of numerical values
# It will threshold this data based on Kapur's entropy method
np.random.seed(42)
data = np.random.randint(0, 256, size=100).astype(np.float64)

# Define ARGS dictionary with all parameters
ARGS = {
    'data': data,
    'step': 255
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
accuracy_assessment Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/acc_ass.py
code served (permissive licence) · get_code("5833dd06a7883c06")
How it ran

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

import numpy as np

np.random.seed(0)

# Create small test arrays for the accuracy assessment function
# The function expects 2D arrays that will be flattened
height, width = 4, 3

# Create ground truth changed map (255 indicates changed pixels)
gt_changed = np.zeros((height, width), dtype=np.uint8)
gt_changed[0:2, 0:2] = 255  # Mark some pixels as changed in ground truth

# Create ground truth unchanged map (255 indicates unchanged pixels)
gt_unchanged = np.zeros((height, width), dtype=np.uint8)
gt_unchanged[2:4, 1:3] = 255  # Mark some pixels as unchanged in ground truth

# Create predicted changed map (255 indicates predicted changed, other values for unchanged)
changed_map = np.zeros((height, width), dtype=np.uint8)
changed_map[0:2, 0:2] = 255  # Predict some pixels as changed
changed_map[2:4, 1:3] = 1    # Predict some pixels as unchanged (non-255 value)

ARGS = {
    "gt_changed": gt_changed,
    "gt_unchanged": gt_unchanged,
    "changed_map": changed_map,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, 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
covw Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/MAD/covw.py
code served (permissive licence) · get_code("443bb565c97c62ea")
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 (5, 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 numpy as np

# Construct small numpy arrays for the covw function
# The function expects:
# - center_X: array that will be concatenated with center_Y along axis 0
# - center_Y: array that will be concatenated with center_X along axis 0
# - w: weight array with shape[1] = n, used for sqrt and sum operations

# Create center_X with shape (3, 4) - 3 samples, 4 features
center_X = np.random.randn(3, 4).astype(np.float64)

# Create center_Y with shape (2, 4) - 2 samples, 4 features
center_Y = np.random.randn(2, 4).astype(np.float64)

# Create weights w with shape (5, 4) where shape[1] = 4 = n
# After concatenation, V will have shape (5, 4), so w should have compatible shape
w = np.abs(np.random.randn(5, 4)).astype(np.float64) + 0.1

ARGS = {
    "center_X": center_X,
    "center_Y": center_Y,
    "w": w
}
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 (5, 5) — shape (5, 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
diff_image Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/PCAKmeans/util.py
code served (permissive licence) · get_code("1caf1ed509ed321f")
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 (4, 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 numpy as np

# Create small test images
image_before = np.random.rand(4, 3, 3).astype(np.float32) * 255
image_after = np.random.rand(4, 3, 3).astype(np.float32) * 255

ARGS = {
    "image_before": image_before,
    "image_after": image_after,
    "is_abs": True,
    "is_multi_channel": False,
}
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 3, dtype family int, shape (4, 3, 3) — shape (4, 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
fuzzy_c_means Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/cluster_util.py
code served (permissive licence) · get_code("5e0b285f13d33951")
How it ran

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

import numpy as np

# Set random seed for reproducibility
np.random.seed(42)

# Create small test data with shape (1, data_item)
# The function expects data with 1 row and multiple columns (data items)
data = np.random.rand(1, 100).astype(np.float64)

ARGS = {
    "data": data,
    "m": 2,
    "types": 4,
    "max_iter": 100,
    "epsilon": 1e-06
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
gene_change_map Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/PCAKmeans/algorithm.py
code served (permissive licence) · get_code("1b1afadf3b455674")
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 (4, 6) 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 stand-in data for the function
# image_diff: 2D array of shape (height, width) representing difference image
img_height, img_width = 4, 6
image_diff = np.random.randint(0, 256, size=(img_height, img_width), dtype=np.uint8).astype(float)

# feature_vectors: 2D array of shape (num_features, num_pixels)
# The function accesses feature_vectors - cluster_center[:, 0] and cluster_center[:, 1]
# So feature_vectors should have shape (num_features, num_pixels) where num_pixels = img_height * img_width
num_features = 5
num_pixels = img_height * img_width
feature_vectors = np.random.randn(num_features, num_pixels)

# cluster_center: 2D array of shape (num_features, 2) - two cluster centers
# The function accesses cluster_center[:, 0] and cluster_center[:, 1]
cluster_center = np.random.randn(num_features, 2)

ARGS = {
    "image_diff": image_diff,
    "feature_vectors": feature_vectors,
    "cluster_center": cluster_center,
}
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 (4, 6) — shape (4, 6) 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
k_means_cluster Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/PCAKmeans/k_means.py
code served (permissive licence) · get_code("3b37ae4cf724aa7a")
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 (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 numpy as np

# Analyze the function:
# - vectors: 2D numpy array with shape (vec_dim, vec_n)
#   Used with: .shape, slicing [:, i], [:, vec_class[0, :] == i]
#   np.linalg.norm and np.mean operations
# - cluster_num: integer
#   Used in: range(), array initialization

# Create small test data
np.random.seed(0)

# Create vectors: (vec_dim=4, vec_n=6)
# This gives us 4-dimensional vectors with 6 samples
vec_dim = 4
vec_n = 6
vectors = np.random.randn(vec_dim, vec_n)

# Number of clusters
cluster_num = 3

ARGS = {
    'vectors': vectors,
    'cluster_num': cluster_num
}
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 (4, 3) — shape (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
norm_img Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/data_prepro.py
code served (permissive licence) · get_code("8118e500fc9b39a6")
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 float, shape (3, 4, 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 numpy as np

# Test case 1: channel_first=True
# Create a small image with shape (3, 4, 5) - 3 channels, 4 height, 5 width
img_channel_first = np.random.rand(3, 4, 5).astype(np.float32) * 255

# Test case 2: channel_first=False
# Create a small image with shape (4, 5, 3) - 4 height, 5 width, 3 channels
img_channel_last = np.random.rand(4, 5, 3).astype(np.float32) * 255

ARGS = {
    "img": img_channel_first,
    "channel_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 ndarray, ndim 3, dtype family float, shape (3, 4, 5) — shape (3, 4, 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
otsu Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/cluster_util.py
code served (permissive licence) · get_code("4510ab89f02a3fce")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float64, ndim 0, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

np.random.seed(0)

# Create sample cluster data
# The function expects data as a 2D array where it accesses data.shape[1]
# and performs operations like data <= value and data > value
data = np.random.rand(10, 20) * 100  # 10x20 array with values 0-100

ARGS = {
    "data": data,
    "num": 400,
    "get_bcm": False,
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
stad_img Ran
ran on a synthesized input
I-Hope-Peace/ChangeDetectionRepository/Methodology/util/data_prepro.py
code served (permissive licence) · get_code("6354fe3264b85dc9")
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 float, shape (3, 4, 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 numpy as np

# Create a small test image with channel_first format
# Shape: (C, H, W) = (3, 4, 5) for distinct dimensions
np.random.seed(0)
img_channel_first = np.random.randn(3, 4, 5).astype(np.float32)

# Create a small test image with channel_last format
# Shape: (H, W, C) = (4, 5, 3) for distinct dimensions
img_channel_last = np.random.randn(4, 5, 3).astype(np.float32)

ARGS = {
    'img': img_channel_first,
    'channel_first': True,
    'get_para': False
}
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 3, dtype family float, shape (3, 4, 5) — shape (3, 4, 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
get_binary_change_map Not yet run
unverified (harvested, no recorded run)
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/MAD/irmad.py
code served (permissive licence) · get_code("9645b670443906b0")
How it ran

no invocation recorded.

zero_pad Not yet run
unverified (harvested, no recorded run)
I-Hope-Peace/ChangeDetectionRepository/Methodology/Traditional/PCAKmeans/util.py
code served (permissive licence) · get_code("3bd0e27058549fe6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 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 12 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

With the development of Earth observation technology, very-high-resolution (VHR) image has become an important data source of change detection. Nowadays, deep learning methods have achieved conspicuous performance in the change detection of VHR images. Nonetheless, most of the existing change detection models based on deep learning require annotated training samples. In this paper, a novel unsupervised model called kernel principal component analysis (KPCA) convolution is proposed for extracting representative features from multi-temporal VHR images. Based on the KPCA convolution, an unsupervised deep siamese KPCA convolutional mapping network (KPCA-MNet) is designed for binary and multi-class change detection. In the KPCA-MNet, the high-level spatial-spectral feature maps are extracted by a deep siamese network consisting of weight-shared PCA convolution layers. Then, the change information in the feature difference map is mapped into a 2-D polar domain. Finally, the change detection results are generated by threshold segmentation and clustering algorithms. All procedures of KPCA-MNet does not require labeled data. The theoretical analysis and experimental results demonstrate the validity, robustness, and potential of the proposed method in two binary change detection data sets and one multi-class change detection data set.

For agents

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

get_harvested_code_for_paper("1912.08628")
get_code_for_paper("1912.08628")
have("1912.08628")

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

curl https://syntology.ai/api/ran/1912.08628.json

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

[![Syntology run record](https://syntology.ai/api/ran/1912.08628.svg)](https://syntology.ai/paper/1912.08628)

Connect an agent — have() is free.