Search for: All records

Creators/Authors contains: "Ni, Zhen"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Federated learning (FL) enables collaborative model trainingwithout centralizing data. In multi-domain scenarios withnon-identically and independently distributed (non-IID) data,prediction performance is often hindered by catastrophic forgettingof specialized local knowledge and negative transferfrom conflicting client updates. To address these challenges,we propose a personalized FL framework with dual-branch(pFedDB) structure and a two-phase training protocol. Thedual-branch architecture separates the model into a sharedbranch for cross-client aggregation and a private branch thatremains on each local client. The private branch is never overwrittenby server updates, which prevents the catastrophicforgetting of domain-specific knowledge. This structure alsosignificantly reduces communication overhead per round asonly the shared branch is transmitted. To mitigate negativetransfer, our two-phase protocol first establishes a personalizedknowledge anchor by training a single-branch expertmodel on each client’s local data. In the second phase, the locallytrained model is cloned to initialize private and sharedbranches. Only the shared branch is aggregated in federatedtraining. This process enables the shared branch to learn ageneral representation that complements the established localexpertise. This design consistently improves the performanceof every client over its single-domain baseline, overcomingthe challenge of negative transfer among clients. Experimentson our new Chest-X-Ray-4 suite and three public benchmarksshow that the proposed pFedDB method obtains 30% savingin communication overhead per round and competitive or betteraccuracy performance than recent FL methods. 
    more » « less
    Free, publicly-accessible full text available March 17, 2027
  2. Free, publicly-accessible full text available January 20, 2027
  3. Free, publicly-accessible full text available December 1, 2026
  4. Schwartz, Peter J; Hohil, Myron E (Ed.)
  5. Mobile robot navigation is a critical aspect of robotics, with applications spanning from service robots to industrial automation. However, navigating in complex and dynamic environments poses many challenges, such as avoiding obstacles, making decisions in real-time, and adapting to new situations. Reinforcement Learning (RL) has emerged as a promising approach to enable robots to learn navigation policies from their interactions with the environment. However, application of RL methods to real-world tasks such as mobile robot navigation, and evaluating their performance under various training–testing settings has not been sufficiently researched. In this paper, we have designed an evaluation framework that investigates the RL algorithm’s generalization capability in regard to unseen scenarios in terms of learning convergence and success rates by transferring learned policies in simulation to physical environments. To achieve this, we designed a simulated environment in Gazebo for training the robot over a high number of episodes. The training environment closely mimics the typical indoor scenarios that a mobile robot can encounter, replicating real-world challenges. For evaluation, we designed physical environments with and without unforeseen indoor scenarios. This evaluation framework outputs statistical metrics, which we then use to conduct an extensive study on a deep RL method, namely the proximal policy optimization (PPO). The results provide valuable insights into the strengths and limitations of the method for mobile robot navigation. Our experiments demonstrate that the trained model from simulations can be deployed to the previously unseen physical world with a success rate of over 88%. The insights gained from our study can assist practitioners and researchers in selecting suitable RL approaches and training–testing settings for their specific robotic navigation tasks. 
    more » « less