Adversarial perturbations reliably fabricate membership signals in vision-model MIAs, separated by a gradient-norm collapse trajectory that enables robust detection and inference.
Cinic-10 is not imagenet or cifar-10
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
abstract
In this brief technical report we introduce the CINIC-10 dataset as a plug-in extended alternative for CIFAR-10. It was compiled by combining CIFAR-10 with images selected and downsampled from the ImageNet database. We present the approach to compiling the dataset, illustrate the example images for different classes, give pixel distributions for each part of the repository, and give some standard benchmarks for well known models. Details for download, usage, and compilation can be found in the associated github repository.
citation-role summary
background 1
dataset 1
citation-polarity summary
years
2026 7representative citing papers
TC-UMIA is a population-level attack using pre- and post-unlearning predictions to infer membership across forget, retain, and unseen sets, revealing added privacy leakage to retained data.
FlashbackCL adds time-decayed label counts, class-balanced replay, and coreset curation to Flashback, yielding 6.9-10% gains and up to 68% less temporal forgetting on CIFAR-10 under controlled shifts.
HARMONY mitigates representation skew in heterogeneous hybrid split federated learning via meta-learning to simulate diverse extractors and server-side contrastive learning to align features, delivering up to 43% accuracy gains.
ArmSSL is a black-box verifiable and adversarially robust watermarking framework for SSL pre-trained encoders using paired discrepancy enlargement, latent entanglement, distribution alignment, and reference-guided tuning.
PSS-MIA with Loss-Gap Ranking pre-selects informative samples for black-box MIAs, outperforming baselines while saving 60-83% of queries under 0.1% FPR on CIFAR-10/100 and CINIC-10.
EvoCSFL combines candidate generation, a multi-objective metric, surrogate approximation, and evolutionary search to optimize client subsets in federated learning, reporting faster convergence and lower energy on image classification tasks.
citing papers explorer
-
A Unified Perspective on Adversarial Membership Manipulation in Vision Models
Adversarial perturbations reliably fabricate membership signals in vision-model MIAs, separated by a gradient-norm collapse trajectory that enables robust detection and inference.
-
Revisiting Privacy Leakage in Machine Unlearning: Membership Inference Beyond the Forgotten Set
TC-UMIA is a population-level attack using pre- and post-unlearning predictions to infer membership across forget, retain, and unseen sets, revealing added privacy leakage to retained data.
-
FlashbackCL: Mitigating Temporal Forgetting in Federated Learning
FlashbackCL adds time-decayed label counts, class-balanced replay, and coreset curation to Flashback, yielding 6.9-10% gains and up to 68% less temporal forgetting on CIFAR-10 under controlled shifts.
-
HARMONY: Bridging the Personalization-Generalization Gap by Mitigating Representation Skew in Heterogeneous Split Federated Learning
HARMONY mitigates representation skew in heterogeneous hybrid split federated learning via meta-learning to simulate diverse extractors and server-side contrastive learning to align features, delivering up to 43% accuracy gains.
-
ArmSSL: Adversarial Robust Black-Box Watermarking for Self-Supervised Learning Pre-trained Encoders
ArmSSL is a black-box verifiable and adversarially robust watermarking framework for SSL pre-trained encoders using paired discrepancy enlargement, latent entanglement, distribution alignment, and reference-guided tuning.
-
Discard the Dross and Select the Essential: Pre-query Sample Selection for Black-box Membership Inference Attacks
PSS-MIA with Loss-Gap Ranking pre-selects informative samples for black-box MIAs, outperforming baselines while saving 60-83% of queries under 0.1% FPR on CIFAR-10/100 and CINIC-10.
-
EvoCSFL: Surrogate-Assisted Evolutionary Client Selection for Efficient and Robust Federated Learning
EvoCSFL combines candidate generation, a multi-objective metric, surrogate approximation, and evolutionary search to optimize client subsets in federated learning, reporting faster convergence and lower energy on image classification tasks.