SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.18133 · CVPR · 2024

Classes Are Not Equal: An Empirical Study on Image Recognition Fairness

Hanwang Zhang, Xiaojuan Qi, Xin Wen, Jiequan Cui, Bei Yu, Beier Zhu

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; 2 honoured the contract we drafted, 0 violated it and 0 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/2402.18133.json · badge: /api/ran/2402.18133.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
dvlab-research/parametric-contrastive-learning canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dvlab-research/parametric-contrastive-learning/GPaCo/LT/paco_imagenet.py
pointer only (licence: MIT); the stored copy get_code serves, from skyz215/proco, is pointer only (licence: NONE) · get_code("b793b278a771fd52")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from skyz215/proco, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

compute_mce Ran
ran, honoured the contract we drafted
dvlab-research/parametric-contrastive-learning/GPaCo/LT/paco_imagenet.py
code served (permissive licence) · get_code("b4e3904ffc1e943d")
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

CORRUPTIONS = ['gaussian_noise', 'shot_noise', 'impulse_noise', 'defocus_blur', 'glass_blur', 'motion_blur', 'zoom_blur', 'snow', 'frost', 'fog', 'brightness', 'contrast', 'elastic_transform', 'pixelate', 'jpeg_compression']

ALEXNET_ERR = [0.886428, 0.894468, 0.92264, 0.81988, 0.826268, 0.785948, 0.79836, 0.866816, 0.826572, 0.819324, 0.564592, 0.853204, 0.646056, 0.71784, 0.6065]

# Create corruption_accs dict with accuracies for each corruption type
# Each corruption maps to an array of accuracies across severity levels
corruption_accs = {}
for corruption in CORRUPTIONS:
    # Generate random accuracies between 0.7 and 0.95 for each corruption
    # Using 5 severity levels as a standard
    corruption_accs[corruption] = np.array([0.75, 0.78, 0.80, 0.82, 0.85])

ARGS = {
    'corruption_accs': corruption_accs
}
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
validate Ran
ran, honoured the contract we drafted
dvlab-research/parametric-contrastive-learning/GPaCo/LT/paco_imagenet.py
code served (permissive licence) · get_code("13eac6ce75be2d11")
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 Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Build a minimal model
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 10)
    
    def forward(self, x):
        return self.linear(x)

# Build a minimal criterion
criterion = nn.CrossEntropyLoss()

# Build minimal data loaders (list of (images, target) tuples)
val_loader = [
    (torch.randn(2, 4), torch.randint(0, 10, (2,))),
    (torch.randn(2, 4), torch.randint(0, 10, (2,))),
]

train_loader = [
    (torch.randn(2, 4), torch.randint(0, 10, (2,))),
    (torch.randn(2, 4), torch.randint(0, 10, (2,))),
]

model = MinimalModel()

# Build minimal args object
class Args:
    def __init__(self):
        self.gpu = None
        self.print_freq = 1

args = Args()

ARGS = {
    'val_loader': val_loader,
    'train_loader': train_loader,
    'model': model,
    'criterion': criterion,
    'args': args,
}
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 Tensor, ndim 0, dtype family float, shape () — shape () 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
test Not yet run
unverified (harvested, no recorded run)
dvlab-research/parametric-contrastive-learning/GPaCo/MAE-ViTs/engine_finetune_paco.py
code served (permissive licence) · get_code("a27000d4fe943463")
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, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 0 later) and 2 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
accuracy 2 other papers
2 dated earlier
2101.10633 (2021-01)
ResLT: Residual Learning for Long-tailed Recognition
2209.00183 (2022-09)
ProCo: Prototype-aware Contrastive Learning for Long-tailed Medical Image Classification
compute_mce 1 other paper
1 dated earlier
2311.14062 (2023-11)
Hardware Resilience Properties of Text-Guided Image Classifiers

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 this paper, we present an empirical study on image recognition unfairness, i.e., extreme class accuracy disparity on balanced data like ImageNet. We demonstrate that classes are not equal and unfairness is prevalent for image classification models across various datasets, network architectures, and model capacities. Moreover, several intriguing properties of fairness are identified. First, the unfairness lies in problematic representation rather than classifier bias distinguished from long-tailed recognition. Second, with the proposed concept of Model Prediction Bias, we investigate the origins of problematic representation during training optimization. Our findings reveal that models tend to exhibit greater prediction biases for classes that are more challenging to recognize. It means that more other classes will be confused with harder classes. Then the False Positives (FPs) will dominate the learning in optimization, thus leading to their poor accuracy. Further, we conclude that data augmentation and representation learning algorithms improve overall performance by promoting fairness to some degree in image classification.

For agents

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

get_harvested_code_for_paper("2402.18133")
get_code_for_paper("2402.18133")
have("2402.18133")

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

curl https://syntology.ai/api/ran/2402.18133.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.18133.svg)](https://syntology.ai/paper/2402.18133)

Connect an agent — have() is free.