arXiv subjects

Peilin He

Publications and source records attributed to Peilin He.

7 recordsLinked to original sources

arXiv2025

Imaging the Topology of Dynamic Brain Connectivity

Functional brain connectivity changes dynamically over time, making its representation challenging for learning on non-Euclidean data. We present a framework that encodes dynamic functional connectivity as an image representation of evolving network topology. Persistent graph homology summarizes global organization across scales, yielding Wasserstein distance-preserving embeddings stable under resolution changes. Stacking these embeddings forms a topological image that captures temporal reconfiguration of brain networks. This design enables convolutional architectures and transfer learning from pretrained foundational models to operate effectively under limited and imbalanced data. Applied to early Alzheimer's detection, the approach achieves clinically meaningful accuracy, establishing a principled foundation for imaging dynamic brain topology.

q-bio.NC↗

arXiv2025

Connectome-Guided Automatic Learning Rates for Deep Networks

The human brain is highly adaptive: its functional connectivity reconfigures on multiple timescales during cognition and learning, enabling flexible information processing. By contrast, artificial neural networks typically rely on manually-tuned learning-rate schedules or generic adaptive optimizers whose hyperparameters remain largely agnostic to a model's internal dynamics. In this paper, we propose Connectome-Guided Automatic Learning Rate (CG-ALR) that dynamically constructs a functional connectome of the neural network from neuron co-activations at each training iteration and adjusts learning rates online as this connectome reconfigures. This connectomics-inspired mechanism adapts step sizes to the network's dynamic functional organization, slowing learning during unstable reconfiguration and accelerating it when stable organization emerges. Our results demonstrate that principles inspired by brain connectomes can inform the design of adaptive learning rates in deep learning, generally outperforming traditional SGD-based schedules and recent methods.

cs.NE↗

arXiv2025

Rigid Invariant Sliced Wasserstein via Independent Embeddings

Comparing probability measures modulo unknown rigid transformations is a central challenge in geometric data analysis. Classical optimal transport (OT) distances, including Wasserstein and sliced Wasserstein, are sensitive to rotations and reflections, whereas Gromov-Wasserstein (GW) and Procrustes-Wasserstein (PW) distances are invariant to isometries but computationally prohibitive for large datasets. We introduce Rigid-Invariant Sliced Wasserstein via Independent Embeddings (RISWIE), a scalable distance that combines the invariance of NP-hard approaches with the efficiency of projection-based OT. RISWIE utilizes data-adaptive bases and matches optimal signed permutations along axes according to distributional similarity to achieve rigid invariance with nearly linear complexity in the sample size. We prove bounds relating RISWIE to GW in special cases and demonstrate dimension-independent statistical stability. Our experiments on cellular imaging and 3D human meshes demonstrate that RISWIE outperforms GW and PW in clustering tasks and discriminative capability while significantly reducing runtime.

cs.CG↗

arXiv2025

Data-Efficient Neural Training with Dynamic Connectomes

The study of dynamic functional connectomes has provided valuable insights into how patterns of brain activity change over time. Neural networks process information through artificial neurons, conceptually inspired by patterns of activation in the brain. However, their hierarchical structure and high-dimensional parameter space pose challenges for understanding and controlling training dynamics. In this study, we introduce a novel approach to characterize training dynamics in neural networks by representing evolving neural activations as functional connectomes and extracting dynamic signatures of activity throughout training. Our results show that these signatures effectively capture key transitions in the functional organization of the network. Building on this analysis, we propose the use of a time series of functional connectomes as an intrinsic indicator of learning progress, enabling a principled early stopping criterion. Our framework performs robustly across benchmarks and provides new insights into neural network training dynamics.

q-bio.NC↗

arXiv2025

PPFL-RDSN: Privacy-Preserving Federated Learning-based Residual Dense Spatial Networks for Encrypted Lossy Image Reconstruction

Reconstructing high-quality images from low-resolution inputs using Residual Dense Spatial Networks (RDSNs) is crucial yet challenging. It is even more challenging in centralized training where multiple collaborating parties are involved, as it poses significant privacy risks, including data leakage and inference attacks, as well as high computational and communication costs. We propose a novel Privacy-Preserving Federated Learning-based RDSN (PPFL-RDSN) framework specifically tailored for encrypted lossy image reconstruction. PPFL-RDSN integrates Federated Learning (FL), local differential privacy, and robust model watermarking techniques to ensure that data remains secure on local clients/devices, safeguards privacy-sensitive information, and maintains model authenticity without revealing underlying data. Empirical evaluations show that PPFL-RDSN achieves comparable performance to the state-of-the-art centralized methods while reducing computational burdens, and effectively mitigates security and privacy vulnerabilities, making it a practical solution for secure and privacy-preserving collaborative computer vision applications.

cs.LG↗

arXiv2025

Conformal-DP: A Density-Aware Mechanism for Differential Privacy over Riemannian Manifolds via Conformal Transformation

Differential Privacy (DP) is being increasingly adopted for non-Euclidean data that lie on complex, high-dimensional manifolds. Existing DP mechanisms for manifold data consider geometric properties when calibrating privacy perturbations, but they largely fail to capture variations in data density within datasets, leading to biased perturbations and suboptimal privacy-utility trade-offs due to heterogeneous data distributions. In this paper, we propose a novel density-aware differential privacy mechanism on Riemannian manifolds, referred to as Conformal-DP, that leverages conformal transformations to calibrate perturbations based on local densities and to induce a density-balanced geometry. We prove that our mechanism satisfies $\epsilon$-differential privacy on any complete Riemannian manifold under mild regularity assumptions. In addition, we derive a closed-form expected geodesic error bound that depends only on the underlying data density ratio and is independent of global curvature. Our empirical results on synthetic and real-world datasets demonstrate that the proposed Conformal-DP mechanism substantially improves the privacy-utility trade-off in heterogeneous data distribution settings, with worst-case performance comparable to state-of-the-art manifold DP mechanisms that assume uniformly distributed data.

cs.CR↗

arXiv2024

DPFedBank: Crafting a Privacy-Preserving Federated Learning Framework for Financial Institutions with Policy Pillars

In recent years, the financial sector has faced growing pressure to adopt advanced machine learning models to derive valuable insights while preserving data privacy. However, the highly sensitive nature of financial data presents significant challenges to sharing and collaboration. This paper presents DPFedBank, an innovative framework enabling financial institutions to collaboratively develop machine learning models while ensuring robust data privacy through Local Differential Privacy (LDP) mechanisms. DPFedBank is designed to address the unique privacy and security challenges associated with financial data, allowing institutions to share insights without exposing sensitive information. By leveraging LDP, the framework ensures that data remains confidential even during collaborative processes, providing a crucial solution for privacy-aware machine learning in finance. We conducted an in-depth evaluation of the potential vulnerabilities within this framework and developed a comprehensive set of policies aimed at mitigating these risks. The proposed policies effectively address threats posed by malicious clients, compromised servers, inherent weaknesses in existing Differential Privacy-Federated Learning (DP-FL) frameworks, and sophisticated external adversaries. Unlike existing DP-FL approaches, DPFedBank introduces a novel combination of adaptive LDP mechanisms and advanced cryptographic techniques specifically tailored for financial data, which significantly enhances privacy while maintaining model utility. Key security enhancements include the implementation of advanced authentication protocols, encryption techniques for secure data exchange, and continuous monitoring systems to detect and respond to malicious activities in real-time.

cs.CE↗