SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2110.01705 · 2021

Let there be a clock on the beach: Reducing Object Hallucination in Image Captioning

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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 0 have no contract verdict; 3 failures of our instrument; 0 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/2110.01705.json · badge: /api/ran/2110.01705.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
furkanbiten/object-bias canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
combine_coco_captions Ran
ran; our contract draft was wrong, not the code
furkanbiten/object-bias/chair/utils/chair.py
code served (permissive licence) · get_code("ee0eaa40d818b077")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 os
import json
import tempfile
import shutil

# Create a temporary directory to hold mock COCO caption files
temp_dir = tempfile.mkdtemp()

# Create mock val captions
val_caps = {
    'info': {'description': 'COCO 2014 Dataset', 'version': '1.0'},
    'licenses': [{'id': 1, 'name': 'License 1'}],
    'images': [
        {'id': 1, 'file_name': 'val_image1.jpg'},
        {'id': 2, 'file_name': 'val_image2.jpg'}
    ],
    'annotations': [
        {'id': 1, 'image_id': 1, 'caption': 'A val caption 1'},
        {'id': 2, 'image_id': 2, 'caption': 'A val caption 2'}
    ]
}

# Create mock train captions
train_caps = {
    'info': {'description': 'COCO 2014 Dataset', 'version': '1.0'},
    'licenses': [{'id': 1, 'name': 'License 1'}],
    'images': [
        {'id': 3, 'file_name': 'train_image1.jpg'},
        {'id': 4, 'file_name': 'train_image2.jpg'}
    ],
    'annotations': [
        {'id': 3, 'image_id': 3, 'caption': 'A train caption 1'},
        {'id': 4, 'image_id': 4, 'caption': 'A train caption 2'}
    ]
}

# Write the mock files
with open(os.path.join(temp_dir, 'captions_val2014.json'), 'w') as f:
    json.dump(val_caps, f)

with open(os.path.join(temp_dir, 'captions_train2014.json'), 'w') as f:
    json.dump(train_caps, f)

ARGS = {
    'annotation_path': temp_dir
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
none outside the standard library — 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
combine_coco_instances Ran
ran; our contract draft was wrong, not the code
furkanbiten/object-bias/chair/utils/chair.py
code served (permissive licence) · get_code("d895aa09a553295e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 os
import tempfile
import shutil

# Create a temporary directory to hold mock annotation files
temp_dir = tempfile.mkdtemp()

# Create mock COCO instance annotations for val set
val_instances = {
    'info': {'description': 'COCO 2014 Val Dataset', 'version': '1.0'},
    'licenses': [{'id': 1, 'name': 'License 1'}],
    'type': 'instances',
    'categories': [
        {'id': 1, 'name': 'person', 'supercategory': 'person'},
        {'id': 2, 'name': 'car', 'supercategory': 'vehicle'}
    ],
    'images': [
        {'id': 1, 'file_name': 'val_image1.jpg', 'width': 640, 'height': 480},
        {'id': 2, 'file_name': 'val_image2.jpg', 'width': 640, 'height': 480}
    ],
    'annotations': [
        {'id': 1, 'image_id': 1, 'category_id': 1, 'bbox': [100, 100, 50, 50], 'area': 2500},
        {'id': 2, 'image_id': 2, 'category_id': 2, 'bbox': [200, 200, 60, 60], 'area': 3600}
    ]
}

# Create mock COCO instance annotations for train set
train_instances = {
    'info': {'description': 'COCO 2014 Train Dataset', 'version': '1.0'},
    'licenses': [{'id': 1, 'name': 'License 1'}],
    'type': 'instances',
    'categories': [
        {'id': 1, 'name': 'person', 'supercategory': 'person'},
        {'id': 2, 'name': 'car', 'supercategory': 'vehicle'}
    ],
    'images': [
        {'id': 3, 'file_name': 'train_image1.jpg', 'width': 640, 'height': 480},
        {'id': 4, 'file_name': 'train_image2.jpg', 'width': 640, 'height': 480}
    ],
    'annotations': [
        {'id': 3, 'image_id': 3, 'category_id': 1, 'bbox': [150, 150, 70, 70], 'area': 4900},
        {'id': 4, 'image_id': 4, 'category_id': 2, 'bbox': [250, 250, 80, 80], 'area': 6400}
    ]
}

# Write the mock annotation files
val_path = os.path.join(temp_dir, 'instances_val2014.json')
train_path = os.path.join(temp_dir, 'instances_train2014.json')

with open(val_path, 'w') as f:
    json.dump(val_instances, f)

with open(train_path, 'w') as f:
    json.dump(train_instances, f)

ARGS = {
    'annotation_path': temp_dir
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
none outside the standard library — 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
load_generated_captions Ran
ran; our contract draft was wrong, not the code
furkanbiten/object-bias/chair/utils/chair.py
code served (permissive licence) · get_code("46d214343afeec61")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 json
import tempfile
import os

# Create a temporary JSON file with the expected structure
caption_data = {
    'overall': {
        'Bleu_1': 0.75,
        'Bleu_2': 0.60,
        'METEOR': 0.25,
        'ROUGE_L': 0.55,
        'CIDEr': 1.0,
        'SPICE': 0.20
    },
    'imgToEval': {
        'img1': {
            'image_id': 'img1',
            'caption': 'A cat sitting on a mat',
            'Bleu_1': 0.8,
            'Bleu_2': 0.65
        },
        'img2': {
            'image_id': 'img2',
            'caption': 'A dog running in the park',
            'Bleu_1': 0.7,
            'Bleu_2': 0.55
        }
    }
}

# Create a temporary file and write the JSON data
temp_file = tempfile.NamedTemporaryFile(mode='w', suffix='.json', delete=False)
json.dump(caption_data, temp_file)
temp_file.close()

ARGS = {
    'cap_file': temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
none outside the standard library — 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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 1 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
combine_coco_captions 10 other papers
10 later
Huang_OPERA_Alleviating_Hallucination_in_Multi-Modal_Large_Language_Models_via_Over-Trust_CVPR_2024_paper (2024 (year only))
OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation
2402.14545 (2024-02)
Less is More: Mitigating Multimodal Hallucination from an EOS Decision Perspective
2405.17220 (2024-05)
RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness
and 7 more not named here
combine_coco_instances 8 other papers
8 later
Huang_OPERA_Alleviating_Hallucination_in_Multi-Modal_Large_Language_Models_via_Over-Trust_CVPR_2024_paper (2024 (year only))
OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation
2402.14545 (2024-02)
Less is More: Mitigating Multimodal Hallucination from an EOS Decision Perspective
2411.16724 (2024-11)
Devils in Middle Layers of Large Vision-Language Models: Interpreting, Detecting and Mitigating Object Hallucinations via Attention Lens
and 5 more not named here

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

Explaining an image with missing or non-existent objects is known as object bias (hallucination) in image captioning. This behaviour is quite common in the state-of-the-art captioning models which is not desirable by humans. To decrease the object hallucination in captioning, we propose three simple yet efficient training augmentation method for sentences which requires no new training data or increase in the model size. By extensive analysis, we show that the proposed methods can significantly diminish our models' object bias on hallucination metrics. Moreover, we experimentally demonstrate that our methods decrease the dependency on the visual features. All of our code, configuration files and model weights will be made public.

For agents

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

get_harvested_code_for_paper("2110.01705")
get_code_for_paper("2110.01705")
have("2110.01705")

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

curl https://syntology.ai/api/ran/2110.01705.json

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

[![Syntology run record](https://syntology.ai/api/ran/2110.01705.svg)](https://syntology.ai/paper/2110.01705)

Connect an agent — have() is free.