SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.01944 · CVPR · 2024

Fourier-basis functions to bridge augmentation gap: Rethinking frequency augmentation in image classification

Shunxin Wang, Nicola Strisciuglio, Puru Vaish

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 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 1 has no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; 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/2403.01944.json · badge: /api/ran/2403.01944.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
nis-research/afa-augment marked official in Papers with Code
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
GeneralFourierOnline Ran
ran on a synthesized input
nis-research/afa-augment/project/augs/fba/fourier_basis.py
code served (permissive licence) · get_code("4975db70c1eb9754")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, 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 torch
import numpy as np
from einops import rearrange
from opt_einsum import contract

class GeneralFourierOnline(torch.nn.Module):

    def __init__(self, img_size, groups, phases, f_cut=1, phase_cut=1, min_str=0, mean_str=5, granularity=64):
        super().__init__()
        _x = torch.linspace(-img_size / 2, img_size / 2, steps=img_size)
        self._x, self._y = torch.meshgrid(_x, _x, indexing='ij')
        self.groups = groups
        self.num_groups = len(groups)
        self.freqs = [f / img_size for f in groups]
        self.phase_range = phases
        self.num_phases = granularity
        self.phases = -np.pi * np.linspace(phases[0], phases[1], num=granularity)
        self.f_cut = f_cut
        self.phase_cut = phase_cut
        self.min_str = min_str
        self.mean_str = mean_str
        self.eps_scale = img_size / 32

    def sample_f_p(self, b, c, device):
        f_cut = self.f_cut
        p_cut = self.phase_cut
        freqs = torch.tensor(self.freqs, device=device, dtype=torch.float32)
        phases = torch.tensor(self.phases, device=device, dtype=torch.float32)
        f_s = freqs[torch.randint(0, self.num_groups, (b, c, f_cut, 1), device=device)]
        p_s = phases[torch.randint(0, self.num_phases, (b, c, f_cut, p_cut), device=device)]
        return (f_s, p_s, f_cut, p_cut)

    def forward(self, x):
        init_shape = x.shape
        if len(x.shape) < 4:
            x = rearrange(x, 'c h w -> () c h w')
        b, c, h, w = x.shape
        freqs, phases, num_f, num_p = self.sample_f_p(b, c, x.device)
        strengths = torch.empty_like(phases).exponential_(1 / self.mean_str) + self.min_str
        return self.apply_fourier_aug(freqs, phases, strengths, x).reshape(init_shape)

    def apply_fourier_aug(self, freqs, phases, strengths, x):
        aug = contract('b c f p, b c f p h w -> b c h w', strengths, self.gen_planar_waves(freqs, phases, x.device))
        aug *= 1 / (self.f_cut * self.phase_cut)
        return torch.clamp(x + aug, 0, 1)

    def gen_planar_waves(self, freqs, phases, device):
        _x, _y = (self._x.to(device), self._y.to(device))
        freqs, phases = (rearrange(freqs, 'b c f p -> b c f p () ()'), rearrange(phases, 'b c f p -> b c f p () ()'))
        _waves = torch.sin(2 * torch.pi * freqs * (_x * torch.cos(phases) + _y * torch.sin(phases)) - torch.pi / 4)
        _waves.div_(_waves.norm(dim=(-2, -1), keepdim=True))
        return self.eps_scale * _waves

    def __str__(self):
        return f'GeneralFourierOnline(f={self.groups}, phases={self.phase_range}, f_cut={self.f_cut}, p_cut={self.phase_cut}, min_str={self.min_str}, max_str={self.mean_str})'


INIT_ARGS = {
    'img_size': 32,
    'groups': [2, 4, 8],
    'phases': [0.0, 1.0],
    'f_cut': 2,
    'phase_cut': 2,
    'min_str': 0.1,
    'mean_str': 3.0,
    'granularity': 16,
}

ARGS = {
    'x': torch.rand(2, 4, 32, 32),
}

if __name__ == '__main__':
    model = GeneralFourierOnline(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
    print(f"Output range: [{output.min():.4f}, {output.max():.4f}]")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
einops, numpy, opt_einsum, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)

Also present in other papers' code

Of 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 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 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.

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

Computer vision models normally witness degraded performance when deployed in real-world scenarios, due to unexpected changes in inputs that were not accounted for during training. Data augmentation is commonly used to address this issue, as it aims to increase data variety and reduce the distribution gap between training and test data. However, common visual augmentations might not guarantee extensive robustness of computer vision models. In this paper, we propose Auxiliary Fourier-basis Augmentation (AFA), a complementary technique targeting augmentation in the frequency domain and filling the robustness gap left by visual augmentations. We demonstrate the utility of augmentation via Fourier-basis additive noise in a straightforward and efficient adversarial setting. Our results show that AFA benefits the robustness of models against common corruptions, OOD generalization, and consistency of performance of models against increasing perturbations, with negligible deficit to the standard performance of models. It can be seamlessly integrated with other augmentation techniques to further boost performance. Codes and models are available at https://github.com/nis-research/afa-augment.

For agents

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

get_harvested_code_for_paper("2403.01944")
get_code_for_paper("2403.01944")
have("2403.01944")

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

curl https://syntology.ai/api/ran/2403.01944.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.01944.svg)](https://syntology.ai/paper/2403.01944)

Connect an agent — have() is free.