MNIST

Dataset Card for MNIST Dataset Summary The MNIST dataset consists of 70,000 28x28 black-and-white images of handwritten digits extracted from two NIST databases. There are 60,000 images in the training dataset and 10,000 images in the validation dataset, one class per digit so a total of 10 classes, with 7,000 images (

Papers using MNIST (27)

Provable Defense Against Privacy Leakage In Federated Learning From Representation Perspective2020 · 156 citesCan Quantum Federated Learning Withstand Circuit-Level Backdoors?2026Data Poisoning Attack Aiming The Vulnerability Of Continual Learning2022 · 2 citesHardening Deep Neural Networks Via Adversarial Model Cascades2018 · 2 citesSemantic Preserving Adversarial Attack Generation With Autoencoder And Genetic Algorithm2022 · 2 citesDetection of Adversarial Examples Through Chaotic Features Extracted From Ordinal Patterns2026Geometrical Perturbations in Generative Models for Black-Box Adversarial Attack2026QShield: Securing Neural Networks Against Adversarial Attacks using Quantum Circuits2026RPP: A Certified Poisoned-Sample Detection Framework for Backdoor Attacks under Dataset Imbalance2026SecureSplit: Mitigating Backdoor Attacks in Split Learning2026SoK: The Last Line of Defense: On Backdoor Defense Evaluation2025SoK: Critical Evaluation of Quantum Machine Learning for Adversarial Robustness2025Integrated Security Mechanisms for Weight Protection in Memristive Crossbar Arrays2025HAMLOCK: HArdware-Model LOgically Combined attacK2025Defending Against Beta Poisoning Attacks in Machine Learning Models2025Robust Federated Learning Under Adversarial Attacks Via Loss-based Client Clustering2025A Little Is Enough: Circumventing Defenses For Distributed Learning2019 · 104 citesAdversarial Ranking Attack And Defense2020 · 32 citesPractical Defences Against Model Inversion Attacks For Split Neural Networks2021 · 26 citesThere Are No Bit Parts For Sign Bits In Black-box Attacks2019 · 14 citesQuery-efficient Adversarial Attack Based On Latin Hypercube Sampling2022 · 11 citesGradient Similarity: An Explainable Approach To Detect Adversarial Attacks Against Deep Learning2018 · 7 citesFedgt: Identification Of Malicious Clients In Federated Learning With Secure Aggregation2023 · 3 citesBroadly Applicable Targeted Data Sample Omission Attacks2021Luring Of Transferable Adversarial Perturbations In The Black-box Paradigm2020Accumulative Poisoning Attacks On Real-time Data2021Client-Side Patching against Backdoor Attacks in Federated Learning2024