SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1807.07428 · 2018

Modeling Visual Context is Key to Augmenting Object Detection Datasets

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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, 1 violated it and 0 have no contract verdict; 2 failures of our instrument; 1 no recorded run), measured 2026-11-21; 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/1807.07428.json · badge: /api/ran/1807.07428.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
dvornikita/context_aug canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
linxi159/context_aug canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
extract_instances_from_image_coco Ran
ran; our contract draft was wrong, not the code
dvornikita/context_aug/Scripts/extract_instances.py
code served (permissive licence) · get_code("35def6338b59c74e")
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)

class MockLoader:
    def __init__(self):
        self.cats = {1, 2, 3}  # Set of valid category IDs
    
    def get_instance_segmentation(self, name):
        # Return masks and categories
        # masks: list of binary masks (each can be a numpy array or tensor)
        # cats: list of category IDs
        masks = [
            np.zeros((4, 5), dtype=np.uint8),
            np.ones((4, 5), dtype=np.uint8),
            np.random.randint(0, 2, (4, 5), dtype=np.uint8),
            np.random.randint(0, 2, (4, 5), dtype=np.uint8)
        ]
        cats = [1, 2, 5, 3]  # Mix of valid (1,2,3) and invalid (5) categories
        return masks, cats

loader = MockLoader()
name = "test_image"

ARGS = {
    "name": name,
    "loader": loader
}
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
get_ends Ran
ran, violated the contract we drafted
dvornikita/context_aug/Scripts/extract_instances.py
code served (permissive licence) · get_code("b796d8564f1d1684")
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 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

# Test case 1: coords with length > 4
coords_long = np.array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0])

# Test case 2: coords with length > 2 but <= 4
coords_medium = np.array([10.0, 20.0, 30.0])

# Test case 3: coords with length <= 2
coords_short = np.array([100.0, 200.0])

# Use the long coords as the primary test case
ARGS = {
    "coords": coords_long
}
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 1, 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
sample2batch Ran
ran; our contract draft was wrong, not the code
linxi159/context_aug/inference.py
code served (permissive licence) · get_code("90b8edfd1317b354")
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

# Construct a sample dict that will be passed to sample2batch
# The function expects:
# - s['bboxes']: array with shape (N, ...) where N is the number of bboxes
# - s['w']: scalar width value
# - s['h']: scalar height value
# - s['img']: image data (will be replicated N times)
# - s['frames']: frame data

N = 2  # Number of bounding boxes

# Create bounding boxes with shape (2, 4) - 2 boxes, 4 coordinates each
bboxes = np.array([[10, 20, 30, 40], [50, 60, 70, 80]], dtype=np.float32)

# Create a small image array (distinct shape to avoid ambiguity)
img = np.random.rand(3, 64, 64).astype(np.float32)

# Create frames data (also distinct shape)
frames = np.array([0, 1], dtype=np.int32)

# Width and height scalars
w = 640
h = 480

s = {
    'bboxes': bboxes,
    'w': w,
    'h': h,
    'img': img,
    'frames': frames
}

ARGS = {'s': s}
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
process_raw_context Not yet run
unverified (harvested, no recorded run)
dvornikita/context_aug/Scripts/encode_context.py
code served (permissive licence) · get_code("636b4eb8e849208c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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 2026-11-21.

Repositories linked to this paper

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

Abstract

Performing data augmentation for learning deep neural networks is well known to be important for training visual recognition systems. By artificially increasing the number of training examples, it helps reducing overfitting and improves generalization. For object detection, classical approaches for data augmentation consist of generating images obtained by basic geometrical transformations and color changes of original training images. In this work, we go one step further and leverage segmentation annotations to increase the number of object instances present on training data. For this approach to be successful, we show that modeling appropriately the visual context surrounding objects is crucial to place them in the right environment. Otherwise, we show that the previous strategy actually hurts. With our context model, we achieve significant mean average precision improvements when few labeled examples are available on the VOC'12 benchmark.

For agents

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

get_harvested_code_for_paper("1807.07428")
get_code_for_paper("1807.07428")
have("1807.07428")

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

curl https://syntology.ai/api/ran/1807.07428.json

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

[![Syntology run record](https://syntology.ai/api/ran/1807.07428.svg)](https://syntology.ai/paper/1807.07428)

Connect an agent — have() is free.