ICLR 2022 accepted papers on Federated learning and privacy
27 of the 1,095 papers accepted at ICLR 2022 (2.5%) are on Federated learning and privacy, compared with 2.0% at ICLR 2021. They fall into 4 topics.
See ICLR 2022 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Personalized and heterogeneous FL · 9 papers
- On Bridging Generic and Personalized Federated Learning for Image Classification (66 citations)
- FedBABU: Toward Enhanced Representation for Federated Image Classification (49 citations)
- Divergence-aware Federated Self-Supervised Learning (33 citations)
- What Do We Mean by Generalization in Federated Learning? (12 citations)
- Federated Learning from Only Unlabeled Data with Class-conditional-sharing Clients (6 citations)
- Towards Model Agnostic Federated Learning Using Knowledge Distillation (4 citations)
- Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization (1 citation)
- An Agnostic Approach to Federated Learning with Class Imbalance
- Diurnal or Nocturnal? Federated Learning of Multi-branch Networks from Periodically Shifting Distributions
Communication-efficient FL · 8 papers
- FedPara: Low-rank Hadamard Product for Communication-Efficient Federated Learning (16 citations)
- Acceleration of Federated Learning with Alleviated Forgetting in Local Training (15 citations)
- Byzantine-Robust Learning on Heterogeneous Datasets via Bucketing (13 citations)
- ZeroFL: Efficient On-Device Training for Federated Learning with Local Sparsity (11 citations)
- FedChain: Chained Algorithms for Near-optimal Communication Cost in Federated Learning (2 citations)
- Recycling Model Updates in Federated Learning: Are Gradient Subspaces Low-Rank? (2 citations)
- Diverse Client Selection for Federated Learning via Submodular Maximization
- Hybrid Local SGD for Federated Learning with Heterogeneous Communications
Differential privacy · 5 papers
- Large Language Models Can Be Strong Differentially Private Learners (72 citations)
- Differentially Private Fine-tuning of Language Models (46 citations)
- Hyperparameter Tuning with Renyi Differential Privacy (6 citations)
- PEARL: Data Synthesis via Private Embeddings and Adversarial Reconstruction Learning (5 citations)
- Differentially Private Fractional Frequency Moments Estimation with Polylogarithmic Space
Privacy attacks and secure learning · 5 papers
- Label Leakage and Protection in Two-party Split Learning (63 citations)
- Bayesian Framework for Gradient Leakage (15 citations)
- Improving Federated Learning Face Recognition via Privacy-Agnostic Clusters (15 citations)
- Robbing the Fed: Directly Obtaining Private Data in Federated Learning with Modified Models (7 citations)
- Privacy Implications of Shuffling
