Anran Li

data heterogeneity (2) non-iid datum (2) federated learning (2) model calibration (1) privacy preservation (1) model aggregation (1) bias mitigation (1) feature extractor (1) global model (1) personalized learning (1) stochastic mutation (1) split federated learning (1) multi-model aggregation (1) knowledge replay (1) multi-modal bayesian inference (1) physical property estimation (1) visual-tactile fusion (1) conjugate pair (1) group contribution (1) gradient-based metric (1)