arXiv subjects

Thanh Tung Nguyen

Publications and source records attributed to Thanh Tung Nguyen.

6 recordsLinked to original sources

arXiv2026

Adjoint-based shape optimization of a ship hull using a Conditional Variational Autoencoder (CVAE) assisted propulsion surrogate model

Adjoint-based shape optimization of ship hulls is a powerful tool for addressing high-dimensional design problems in naval architecture, particularly in minimizing the ship resistance. However, its application to vessels that employ complex propulsion systems introduces significant challenges. They arise from the need for transient simulations extending over long periods of time with small time steps and from the reverse temporal propagation of the primal and adjoint solutions. These challenges place considerable demands on the required storage and computing power, which significantly hamper the use of adjoint methods in the industry. To address this issue, we propose a machine learning-assisted optimization framework that employs a Conditional Variational Autoencoder-based surrogate model of the propulsion system. The surrogate model replicates the time-averaged flow field induced by a Voith Schneider Propeller and replaces the geometrically and time-resolved propeller with a data-driven approximation. Primal flow verification examples demonstrate that the surrogate model achieves significant computational savings while maintaining the necessary accuracy of the resolved propeller. Optimization studies demonstrate that neglecting the propulsion system can result in hull designs whose performance is inferior to that of the initial shape when subsequently validated using a numerically resolved propulsor. In contrast, the proposed method produces shapes that actually achieve more than an 8% reduction in resistance.

physics.flu-dyn↗

arXiv2026

Stability-Aware Imitation Learning from Model Predictive Control for Autonomous Vehicle Lateral Control: Exact Q-Loss and a Novel Training Procedure

This paper develops a certified imitation-learning framework for approximating model predictive control (MPC) policies with feedforward neural controllers and validates it on autonomous-vehicle lateral control. An exact finite-horizon Q-loss is constructed by fixing the learner's first steering action in the expert MPC problem and re-optimizing the remaining horizon, thereby measuring its downstream optimal-control consequence rather than only pointwise action mismatch. The neural policy is represented as a linear fractional transformation (LFT) interconnection with activation nonlinearities described by sector integral quadratic constraints (IQCs). Combined with a quadratic Lyapunov condition, this representation yields a differentiable certification margin based on the largest eigenvalue of the Lyapunov-IQC matrix. The margin is enforced during training through a logarithmic barrier, while certified Dataset Aggregation (DAgger) and safe projection keep data-aggregation rollouts within the certified policy set. Experiments on a CAD-referenced autonomous-vehicle platform with AprilTag localization and real-time steering demonstrate the resulting closed-loop performance.

eess.SY↗

arXiv2022

LAPFormer: A Light and Accurate Polyp Segmentation Transformer

Polyp segmentation is still known as a difficult problem due to the large variety of polyp shapes, scanning and labeling modalities. This prevents deep learning model to generalize well on unseen data. However, Transformer-based approach recently has achieved some remarkable results on performance with the ability of extracting global context better than CNN-based architecture and yet lead to better generalization. To leverage this strength of Transformer, we propose a new model with encoder-decoder architecture named LAPFormer, which uses a hierarchical Transformer encoder to better extract global feature and combine with our novel CNN (Convolutional Neural Network) decoder for capturing local appearance of the polyps. Our proposed decoder contains a progressive feature fusion module designed for fusing feature from upper scales and lower scales and enable multi-scale features to be more correlative. Besides, we also use feature refinement module and feature selection module for processing feature. We test our model on five popular benchmark datasets for polyp segmentation, including Kvasir, CVC-Clinic DB, CVC-ColonDB, CVC-T, and ETIS-Larib

cs.CV↗

arXiv2022

Online pseudo labeling for polyp segmentation with momentum networks

Semantic segmentation is an essential task in developing medical image diagnosis systems. However, building an annotated medical dataset is expensive. Thus, semi-supervised methods are significant in this circumstance. In semi-supervised learning, the quality of labels plays a crucial role in model performance. In this work, we present a new pseudo labeling strategy that enhances the quality of pseudo labels used for training student networks. We follow the multi-stage semi-supervised training approach, which trains a teacher model on a labeled dataset and then uses the trained teacher to render pseudo labels for student training. By doing so, the pseudo labels will be updated and more precise as training progress. The key difference between previous and our methods is that we update the teacher model during the student training process. So the quality of pseudo labels is improved during the student training process. We also propose a simple but effective strategy to enhance the quality of pseudo labels using a momentum model -- a slow copy version of the original model during training. By applying the momentum model combined with re-rendering pseudo labels during student training, we achieved an average of 84.1% Dice Score on five datasets (i.e., Kvarsir, CVC-ClinicDB, ETIS-LaribPolypDB, CVC-ColonDB, and CVC-300) with only 20% of the dataset used as labeled data. Our results surpass common practice by 3% and even approach fully-supervised results on some datasets. Our source code and pre-trained models are available at https://github.com/sun-asterisk-research/online learning ssl

cs.CV↗

arXiv2022

Adjoint Node-Based Shape Optimization of Free Floating Vessels

The paper is concerned with a node-based, gradient-driven, continuous adjoint two-phase flow procedure to optimize the shapes of free-floating vessels and discusses three topics. First, we aim to convey that elements of a Cahn-Hilliard formulation should augment the frequently employed Volume-of-Fluid two-phase flow model to maintain dual consistency. It is seen that such consistency serves as the basis for a robust primal/adjoint coupling in practical applications at huge Reynolds and Froude numbers. The second topic covers different adjoint coupling strategies. A central aspect of the application is the floating position, particularly the trim and the sinkage, that interact with a variation of hydrodynamic loads induced by the shape updates. Other topics addressed refer to the required level of density coupling and a more straightforward -- yet non-frozen -- adjoint treatment of turbulence. The third part discusses the computation of a descent direction within a node-based environment. We will illustrate means to deform both the volume mesh and the hull shape simultaneously and at the same time obey technical constraints on the vessel's displacement and its extensions. The Hilbert-space approach provides smooth shape updates using the established coding infrastructure of a computational fluid dynamics algorithm and provides access to managing additional technical constraints. Verification and validation follow from a submerged 2D cylinder case. The application includes a full-scale offshore supply vessel at Re=3E+08 and Fn=0.37. Results illustrate that the fully parallel procedure can automatically reduce the drag of an already pre-optimized shape by 9-13 within approximately O(10.000-30.000) CPUh depending on the considered couplings and floatation aspects.

physics.flu-dyn↗

arXiv2008

Development of a peristaltic micropump for bio-medical applications based on mini LIPCA

This paper presents the design, fabrication, and experimental characterization of a peristaltic micropump. The micropump is composed of two layers fabricated from polydimethylsiloxane (PDMS) material. The first layer has a rectangular channel and two valve seals. Three rectangular mini lightweight piezo-composite actuators are integrated in the second layer, and used as actuation parts. Two layers are bonded, and covered by two polymethyl methacrylate (PMMA) plates, which help increase the stiffness of the micropump. A maximum flow rate of 900 mokroliter per min and a maximum backpressure of 1.8 kPa are recorded when water is used as pump liquid. We measured the power consumption of the micropump. The micropump is found to be a promising candidate for bio-medical application due to its bio-compatibility, portability, bidirectionality, and simple effective design.

cs.RO↗