# Hyperparticle/one-pixel-attack-keras

Keras implementation of "One pixel attack for fooling deep neural networks" using differential evolution on Cifar10 and ImageNet

Repository: https://github.com/Hyperparticle/one-pixel-attack-keras
Canonical: https://ross.abutalabs.com/products/one-pixel-attack-keras
Homepage: https://arxiv.org/abs/1710.08864
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: keras, cnn, cifar10, machine-learning, tensorflow, deep-learning, neural-network, imagenet, image-processing, nlp
Last push: 2024-04-24T12:00:12+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3119, "days_push": 861, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1237, forks 214 (observed 2026-08-28T04:04:05.414915+00:00)

## What it is
A Keras/TensorFlow implementation and tutorial of the 'One Pixel Attack' paper, which fools deep neural networks by modifying a single pixel using differential evolution. It includes Jupyter notebooks demonstrating black-box adversarial attacks on CIFAR-10 and ImageNet classifiers.

## Use cases
- generate adversarial examples to fool image classifiers
- learn how differential evolution attacks neural networks
- evaluate robustness of a CNN against one-pixel perturbations
- reproduce the one pixel attack paper in Keras
- study black-box adversarial machine learning attacks
- run adversarial attack experiments on CIFAR-10 in Colab

## When to choose
- you want a runnable, educational implementation of the one-pixel attack with notebooks
- you use Keras/TensorFlow and want to test model robustness with a black-box evolutionary attack
- you are studying adversarial machine learning and want to see differential evolution applied to attacks

## When to avoid
- you need production-grade adversarial training or defense tooling
- you use PyTorch or need attacks beyond one-pixel perturbations
- you need actively maintained code for the latest TensorFlow versions

## Facets
- artifact type: learning-resource
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, security
- domain: machine-learning, deep-learning, computer-vision, security
- platform: python
- tags: adversarial-attacks, differential-evolution, keras, tensorflow, cifar10, imagenet, jupyter-notebook, black-box-attack, tutorial, gpu

## Member repositories
- Hyperparticle/one-pixel-attack-keras (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:05.414915+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:21:50.174147+00:00, confidence not recorded.
  - readme: https://github.com/Hyperparticle/one-pixel-attack-keras (fetched 2026-08-28T04:04:05.414915+00:00, sha cebb533e270a)
  - homepage: https://arxiv.org/abs/1710.08864 (fetched 2026-08-29T12:21:17.030068+00:00, sha 4c9dc1b54dce)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:21:17.044331+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:21:17.040199+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:21:17.046673+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:21:17.042436+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.