Research map / Adversarial robustness and security / Certified and empirical robustness
Adversarial Learning for Robust Deep Clustering
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add adversarial defense (perturbation-based contrastive constraints) to the clustering network | ConvAE · MNIST, dim 128 | ACC (%) on clean features: 84.9 -> 87.2 | Helped |
| add adversarial defense | ConvAE+MIE · MNIST, dim 128 | ACC (%) on clean features: 91.3 -> 94.2 | Helped |
| add adversarial defense | ConvAE+Graph · MNIST, dim 128 | ACC (%) on clean features: 96.2 -> 98.1 | Helped |
| add adversarial defense, evaluated under the learned attack | ConvAE · MNIST, dim 128 | ACC (%) on perturbed features: 77.2 -> 87.0 | Helped |
| add adversarial defense, evaluated under the learned attack | ConvAE+Graph · MNIST, dim 128 | ACC (%) on perturbed features: 88.5 -> 97.7 | Helped |
| add adversarial defense | IDEC · MNIST | ACC (%): 87.4 -> 95.2 | Helped |
| add adversarial defense | VaDE · MNIST | ACC (%): 94.3 -> 96.7 | Helped |
| add adversarial defense | DANDSC · Fashion-MNIST | ACC (%): 67.4 -> 68.7 | Helped |
| train defense with learned adversarial perturbation instead of random noise (ablation) | defense strategy · MNIST | robustness and clustering accuracy | Helped |
| vary defense weight lambda in [0.1, 1] (ablation) | ConvAE+ and MIE+ · MNIST | ACC on perturbed features | No clear effect |
| add MIE or graph module to the basic clustering network | ConvAE · MNIST | robustness to perturbation | No clear effect |
Closest papers by meaning
- Adversarially Robust Deep Multi-View Clustering: A Novel Attack and Defense Framework (ICML 2024 · similarity 0.92)
- On the Adversarial Robustness of Multi-Kernel Clustering (ICML 2025 · similarity 0.88)
- Hard Regularization to Prevent Deep Online Clustering Collapse without Data Augmentation (AAAI 2024 · similarity 0.85)
- Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data (ICLR 2025 · similarity 0.85)
- You Can Trust Your Clustering Model: A Parameter-free Self-Boosting Plug-in for Deep Clustering (NeurIPS 2025 · similarity 0.85)
- Learning from Sample Stability for Deep Clustering (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.