Author name cluster

Lui Sha

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

7 papers

2 author rows

TMLR Journal 2026 Journal Article

Weakly-Supervised Disentangled Representation Learning via Filter-Based Adaptive Swapping

Zhenyu Zong
Qidi Wang
Simon Yu
Hongpeng Cao
Yanbing Mao
Han Zhao
Lui Sha
Huajie Shao

Disentangled representation learning (DRL) aims to uncover semantically meaningful latent factors from observed data, thereby improving both interpretability and generalization of machine learning (ML) models. Despite remarkable progress, unsupervised DRL cannot achieve complete disentanglement without inductive biases or supervision. To address this challenge, existing approaches either rely on full supervision, which demands extensive manual labeling, or weak supervision, which involves complex training strategies that often result in unstable training. To address these limitations, we propose Filter-VAE, a weakly supervised variational autoencoder (VAE) that introduces a filter-based adaptive swapping strategy to learn stable and meaningful disentangled representations. Specifically, a relevance filter removes semantically meaningless latent factors, while an adaptive swapping filter exchanges those latent factors that have reached stability. With these two filters, Filter-VAE adaptively swaps only stable and semantically aligned latent factors, leading to robust and meaningful representations. We evaluate Filter-VAE on three standard benchmarks and our created traffic sign dataset in two downstream tasks: disentanglement and adversarial robustness. Experimental results demonstrate that Filter-VAE achieves strong disentanglement performance with reduced supervision and delivers remarkable robustness against diverse adversarial attacks and corruptions. The code is released at https://github.com/ZY-Zong/Filter-VAE.git.

PDF Details

ICLR Conference 2025 Conference Paper

Accelerating Neural ODEs: A Variational Formulation-based Approach

Hongjue Zhao
Yuchen Wang
Hairong Qi 0001
Zijie Huang 0002
Han Zhao 0002
Lui Sha
Huajie Shao

Neural Ordinary Differential Equations (Neural ODEs or NODEs) excel at modeling continuous dynamical systems from observational data, especially when the data is irregularly sampled. However, existing training methods predominantly rely on numerical ODE solvers, which are time-consuming and prone to accumulating numerical errors over time due to autoregression. In this work, we propose VF-NODE, a novel approach based on the variational formulation (VF) to accelerate the training of NODEs. Unlike existing training methods, the proposed VF-NODEs implement a series of global integrals, thus evaluating Deep Neural Network (DNN)--based vector fields only at specific observed data points. This strategy drastically reduces the number of function evaluations (NFEs). Moreover, our method eliminates the use of autoregression, thereby reducing error accumulations for modeling dynamical systems. Nevertheless, the VF loss introduces oscillatory terms into the integrals when using the Fourier basis. We incorporate Filon's method to address this issue. To further enhance the performance for noisy and incomplete data, we employ the natural cubic spline regression to estimate a closed-form approximation. We provide a fundamental analysis of how our approach minimizes computational costs. Extensive experiments demonstrate that our approach accelerates NODE training by 10 to 1000 times compared to existing NODE-based methods, while achieving higher or comparable accuracy in dynamical systems. The code is available at https://github.com/ZhaoHongjue/VF-NODE-ICLR2025.

Details

NeurIPS Conference 2025 Conference Paper

Real-DRL: Teach and Learn at Runtime

Yanbing Mao
Yihao Cai
Lui Sha

This paper introduces the Real-DRL framework for safety-critical autonomous systems, enabling runtime learning of a deep reinforcement learning (DRL) agent to develop safe and high-performance action policies in real plants while prioritizing safety. The Real-DRL consists of three interactive components: a DRL-Student, a PHY-Teacher, and a Trigger. The DRL-Student is a DRL agent that innovates in the dual self-learning and teaching-to-learn paradigm and the safety-status-dependent batch sampling. On the other hand, PHY-Teacher is a physics-model-based design of action policies that focuses solely on safety-critical functions. PHY-Teacher is novel in its real-time patch for two key missions: i) fostering the teaching-to-learn paradigm for DRL-Student and ii) backing up the safety of real plants. The Trigger manages the interaction between the DRL-Student and the PHY-Teacher. Powered by the three interactive components, the Real-DRL can effectively address safety challenges that arise from the unknown unknowns and the Sim2Real gap. Additionally, Real-DRL notably features i) assured safety, ii) automatic hierarchy learning (i. e. , safety-first learning and then high-performance learning), and iii) safety-informed batch sampling to address the experience imbalance caused by corner cases. Experiments with a real quadruped robot, a quadruped robot in Nvidia Isaac Gym, and a cart-pole system, along with comparisons and ablation studies, demonstrate the Real-DRL's effectiveness and unique features.

PDF Details

ICLR Conference 2024 Conference Paper

Physics-Regulated Deep Reinforcement Learning: Invariant Embeddings

Hongpeng Cao
Yanbing Mao
Lui Sha
Marco Caccamo

This paper proposes the Phy-DRL: a physics-regulated deep reinforcement learning (DRL) framework for safety-critical autonomous systems. The Phy-DRL has three distinguished invariant-embedding designs: i) residual action policy (i.e., integrating data-driven-DRL action policy and physics-model-based action policy), ii) automatically constructed safety-embedded reward, and iii) physics-model-guided neural network (NN) editing, including link editing and activation editing. Theoretically, the Phy-DRL exhibits 1) a mathematically provable safety guarantee and 2) strict compliance of critic and actor networks with physics knowledge about the action-value function and action policy. Finally, we evaluate the Phy-DRL on a cart-pole system and a quadruped robot. The experiments validate our theoretical results and demonstrate that Phy-DRL features guaranteed safety compared to purely data-driven DRL and solely model-based design while offering remarkably fewer learning parameters and fast training towards safety guarantee.

Details

AIJ Journal 2023 Journal Article

Robust vehicle lane keeping control with networked proactive adaptation

Hunmin Kim
Wenbin Wan
Naira Hovakimyan
Lui Sha
Petros Voulgaris

Road condition is an important environmental factor for autonomous vehicle control. A dramatic change in the road condition from the nominal status is a source of uncertainty that can lead to a system failure. Once the vehicle encounters an uncertain environment, such as hitting an ice patch, it is too late to reduce the speed, and the vehicle can lose control. To cope with unforeseen uncertainties in advance, we study a proactive robust adaptive control architecture for autonomous vehicles' lane-keeping control problems. In the proposed framework, the data center generates a prior environmental uncertainty estimate with a quantified uncertainty by combining weather forecasts and measurements from anonymous vehicles through a spatio-temporal filter. The prior estimate and quantified uncertainty contribute to designing a robust heading controller and nominal longitudinal velocity for proactive adaptation to each new abnormal condition. Then the control parameters are updated based on posterior information fusion with on-board measurements.

Details DOI

JBHI Journal 2015 Journal Article

The Design of Safe Networked Supervisory Medical Systems Using Organ-Centric Hierarchical Control Architecture

Woochul Kang
Lui Sha
Richard B. Berlin
Julian M. Goldman

There are growing demands to leverage network connectivity and interoperability of medical devices in order to improve patient safety and the effectiveness of medical services. However, if not properly designed, the integration of medical devices through networking could significantly increase the complexity of the system and make the system more vulnerable to potential errors, jeopardizing patient safety. The system must be designed and verified to guarantee the safety of patients and the effectiveness of medical services in the face of potential problems such as network failures. In this paper, we propose organ-centric hierarchical control architecture as a viable solution that reduces the complexity in system design and verification. In our approach, medical devices are grouped into clusters according to organ-specific human physiology. Each cluster captures common patterns arising out of medical device interactions and becomes a survivable semiautonomous unit during network failures. Further, safety verification and runtime enforcement can be modularized along organ-centric hierarchical control structure. We show the feasibility of the proposed approach under Simulink's model-based development framework. A simplified scenario for airway laser surgery is used as a case study.

Details DOI

JBHI Journal 2013 Journal Article

Design and QoS of a Wireless System for Real-Time Remote Electrocardiography

Kyungtae Kang
Junhee Ryu
Junbeom Hur
Lui Sha

Quality of service (QoS) and, in particular, reliability and a bounded low latency are essential attributes of safety-critical wireless systems for medical applications. However, wireless links are typically prone to bursts of errors, with characteristics which vary over time. We propose a wireless system suitable for real-time remote patient monitoring in which the necessary reliability and guaranteed latency are both achieved by an efficient error control scheme. We have paired an example remote electrocardiography application to this wireless system. We also developed a tool chain that uses a formal description of the proposed wireless medical system architecture in the architecture analysis and design language to assess various combinations of system parameters: we can determine the QoS in terms of packet-delivery ratio and the service latency, and also the size of jitter buffer required for seamless ECG monitoring. A realistic assessment, based on data from the MIT-BIT arrhythmia database, shows that the proposed wireless system can achieve an appropriate level of QoS for real-time ECG monitoring if link-level error control is correctly implemented. Additionally, we present guidelines for the design of energy-efficient link-level error control, derived from energy data, obtained from simulations.

Details DOI