SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2211.12979 · 2022

FLAIR #1: semantic segmentation and domain adaptation dataset

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 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 7 have no contract verdict; 1 failure of our instrument; 1 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/2211.12979.json · badge: /api/ran/2211.12979.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
IGNF/FLAIR-1-AI-Challenge canonical (classifier label)
classifier basis not recorded on the link
8 ran · 1 no recorded run
FunctionStatusWhere it lives
class_IoU Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/flair/metrics.py
code served (permissive licence) · get_code("8605ce289d024558")
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

# The function class_IoU takes:
# - npcm: a confusion matrix (2D numpy array)
# - n_class: number of classes (int)

# The function uses npcm with:
# - np.diag(npcm) to get diagonal elements
# - np.sum(npcm, axis=1) to sum rows
# - np.sum(npcm, axis=0) to sum columns
# - np.isnan() check on the result
# - np.mean() on the result

# Build a small confusion matrix for testing
# Let's use 4 classes (to keep it small but distinct)
# Confusion matrix should be square (n_class x n_class)
npcm = np.array([
    [50, 5, 2, 1],
    [3, 40, 4, 2],
    [1, 2, 35, 3],
    [2, 1, 1, 30]
], dtype=np.float32)

n_class = 4

ARGS = {
    "npcm": npcm,
    "n_class": n_class
}
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
class_precision Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/flair/metrics.py
code served (permissive licence) · get_code("c68ca4f8433fcdb5")
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 a small confusion matrix (npcm parameter)
# This should be a 2D numpy array where:
# - rows represent true classes
# - columns represent predicted classes
# The function uses np.diag(npcm) to get diagonal elements
# and np.sum(npcm, axis=0) to sum along axis 0 (columns)

npcm = np.array([
    [50, 5, 2],
    [3, 45, 2],
    [2, 3, 48]
], dtype=np.float32)

ARGS = {
    "npcm": npcm
}
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
convert Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/zone_detect/dataset.py
code served (permissive licence) · get_code("5e081aa42e00bf1e")
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 int, shape (2, 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 numpy as np

# Test case 1: class_prob with values > 1 (integer type)
img_class_prob_int = np.array([[100, 200], [150, 250]], dtype=np.uint16)

# Test case 2: class_prob with values <= 1 (float type)
img_class_prob_float = np.array([[0.2, 0.8], [0.5, 0.9]], dtype=np.float32)

# Test case 3: argmax with shape (2, 4, 3) - 2 classes, 4x3 spatial
img_argmax = np.array([[[0.1, 0.2, 0.3], [0.4, 0.5, 0.6], [0.7, 0.8, 0.9], [0.1, 0.2, 0.3]],
                        [[0.9, 0.8, 0.7], [0.6, 0.5, 0.4], [0.3, 0.2, 0.1], [0.9, 0.8, 0.7]]], dtype=np.float32)

# Create multiple test configurations
ARGS = [
    {'img': img_class_prob_int, 'img_type': 'class_prob'},
    {'img': img_class_prob_float, 'img_type': 'class_prob'},
    {'img': img_argmax, 'img_type': 'argmax'},
    {'img': img_class_prob_int, 'img_type': 'unknown_type'},
]

# Use the first configuration as the primary ARGS dict
ARGS = {'img': img_class_prob_int, 'img_type': 'class_prob'}
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 int, shape (2, 2) — shape (2, 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
gather_paths Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/flair/tasks_utils.py
code served (permissive licence) · get_code("9336d30f0d66ba61")
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 json
import tempfile
import os

# Create temporary directory and files for the test
tmpdir = tempfile.mkdtemp()

# Create a sample CSV file for training split
train_csv_path = os.path.join(tmpdir, 'train.csv')
with open(train_csv_path, 'w') as f:
    f.write('/path/to/image1.jpg,/path/to/label1.jpg\n')
    f.write('/path/to/image2.jpg,/path/to/label2.jpg\n')

# Create a sample metadata JSON file
metadata_path = os.path.join(tmpdir, 'metadata.json')
metadata = {
    'image1': {
        'patch_centroid_x': 1000000.0,
        'patch_centroid_y': 2000000.0,
        'patch_centroid_z': 500.5,
        'camera': 'UCE_camera',
        'date': '2020-06-15',
        'time': '14h30'
    },
    'image2': {
        'patch_centroid_x': 1500000.0,
        'patch_centroid_y': 2500000.0,
        'patch_centroid_z': 1000.0,
        'camera': 'OTHER_camera',
        'date': '2019-03-20',
        'time': '09h45'
    }
}
with open(metadata_path, 'w') as f:
    json.dump(metadata, f)

# Create config dictionary
config = {
    'paths': {
        'train_csv': train_csv_path,
        'val_csv': None,
        'test_csv': None,
        'path_metadata_aerial': metadata_path
    },
    'use_metadata': True
}

ARGS = {
    'config': config,
    'split': 'train'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — 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
get_module Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/zone_detect/model.py
code served (permissive licence) · get_code("b22bf961966c9438")
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 torch
import os
from pathlib import Path
from collections import OrderedDict
import tempfile

torch.manual_seed(0)

# Create a temporary checkpoint file for testing
temp_dir = tempfile.mkdtemp()
checkpoint_path = os.path.join(temp_dir, "model.ckpt")

# Create sample weights with the expected structure
sample_weights = {
    'model.seg_model.layer1.weight': torch.randn(4, 3),
    'model.seg_model.layer1.bias': torch.randn(4),
    'model.seg_model.layer2.weight': torch.randn(5, 4),
}

# Save as .ckpt format (dictionary with 'state_dict' key)
torch.save({'state_dict': sample_weights}, checkpoint_path)

ARGS = {
    'checkpoint': checkpoint_path
}
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 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
norm Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/flair/data_loader.py
code served (permissive licence) · get_code("4e84b594875d4a4e")
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 as numpy array
# Using shape (3, 4, 5) to represent (channels, height, width)
in_img = np.random.randint(0, 256, size=(3, 4, 5), dtype=np.uint8)

# Set up normalization parameters for custom normalization
means = [100.0, 110.0, 120.0]  # One mean per channel
stds = [30.0, 35.0, 40.0]      # One std per channel

ARGS = {
    'in_img': in_img,
    'norm_type': 'custom',
    'means': means,
    'stds': stds
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, skimage — 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
overall_accuracy Ran
ran on a synthesized input
IGNF/FLAIR-1-AI-Challenge/src/flair/metrics.py
code served (permissive licence) · get_code("0ef823e46507c68d")
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

# The function expects a numpy confusion matrix (npcm)
# It calls npcm.sum() and np.trace(npcm), so it needs a 2D square numpy array
# Let's create a small confusion matrix for a 3-class classification problem

npcm = np.array([
    [50, 5, 2],
    [3, 45, 7],
    [2, 8, 48]
], dtype=np.int32)

ARGS = {
    "npcm": npcm
}
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
read_config Ran
ran; our contract draft was wrong, not the code
IGNF/FLAIR-1-AI-Challenge/src/flair/utils.py
code served (permissive licence) · get_code("f99b92265e606433")
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 yaml
import tempfile
import os

# Create a temporary YAML file with valid content
temp_dir = tempfile.mkdtemp()
config_file = os.path.join(temp_dir, "config.yaml")

config_content = """
model:
  name: "test_model"
  layers: 3
  hidden_dim: 128
training:
  batch_size: 32
  learning_rate: 0.001
  epochs: 10
data:
  train_path: "/path/to/train"
  val_path: "/path/to/val"
"""

with open(config_file, 'w') as f:
    f.write(config_content)

ARGS = {
    "file_path": config_file
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
yaml — 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
train Not yet run
unverified (harvested, no recorded run)
IGNF/FLAIR-1-AI-Challenge/src/flair/tasks.py
code served (permissive licence) · get_code("a583fc74263240ea")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 8 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
read_config 1 other paper
1 later
2310.13336 (2023-10)
FLAIR: a Country-Scale Land Cover Semantic Segmentation Dataset From Multi-Source Optical Imagery

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

The French National Institute of Geographical and Forest Information (IGN) has the mission to document and measure land-cover on French territory and provides referential geographical datasets, including high-resolution aerial images and topographic maps. The monitoring of land-cover plays a crucial role in land management and planning initiatives, which can have significant socio-economic and environmental impact. Together with remote sensing technologies, artificial intelligence (IA) promises to become a powerful tool in determining land-cover and its evolution. IGN is currently exploring the potential of IA in the production of high-resolution land cover maps. Notably, deep learning methods are employed to obtain a semantic segmentation of aerial images. However, territories as large as France imply heterogeneous contexts: variations in landscapes and image acquisition make it challenging to provide uniform, reliable and accurate results across all of France. The FLAIR-one dataset presented is part of the dataset currently used at IGN to establish the French national reference land cover map "Occupation du sol à grande échelle" (OCS- GE).

For agents

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

get_harvested_code_for_paper("2211.12979")
get_code_for_paper("2211.12979")
have("2211.12979")

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

curl https://syntology.ai/api/ran/2211.12979.json

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

[![Syntology run record](https://syntology.ai/api/ran/2211.12979.svg)](https://syntology.ai/paper/2211.12979)

Connect an agent — have() is free.