arXiv subjects

Vinh Nguyen

Publications and source records attributed to Vinh Nguyen.

At least 37 records · Page 2Linked to original sources

arXiv2025

Multimodal Object Detection using Depth and Image Data for Manufacturing Parts

Manufacturing requires reliable object detection methods for precise picking and handling of diverse types of manufacturing parts and components. Traditional object detection methods utilize either only 2D images from cameras or 3D data from lidars or similar 3D sensors. However, each of these sensors have weaknesses and limitations. Cameras do not have depth perception and 3D sensors typically do not carry color information. These weaknesses can undermine the reliability and robustness of industrial manufacturing systems. To address these challenges, this work proposes a multi-sensor system combining an red-green-blue (RGB) camera and a 3D point cloud sensor. The two sensors are calibrated for precise alignment of the multimodal data captured from the two hardware devices. A novel multimodal object detection method is developed to process both RGB and depth data. This object detector is based on the Faster R-CNN baseline that was originally designed to process only camera images. The results show that the multimodal model significantly outperforms the depth-only and RGB-only baselines on established object detection metrics. More specifically, the multimodal model improves mAP by 13% and raises Mean Precision by 11.8% in comparison to the RGB-only baseline. Compared to the depth-only baseline, it improves mAP by 78% and raises Mean Precision by 57%. Hence, this method facilitates more reliable and robust object detection in service to smart manufacturing applications.

cs.CV↗

arXiv2025

Sanitizing Manufacturing Dataset Labels Using Vision-Language Models

The success of machine learning models in industrial applications is heavily dependent on the quality of the datasets used to train the models. However, large-scale datasets, specially those constructed from crowd-sourcing and web-scraping, often suffer from label noise, inconsistencies, and errors. This problem is particularly pronounced in manufacturing domains, where obtaining high-quality labels is costly and time-consuming. This paper introduces Vision-Language Sanitization and Refinement (VLSR), which is a vision-language-based framework for label sanitization and refinement in multi-label manufacturing image datasets. This method embeds both images and their associated textual labels into a shared semantic space leveraging the CLIP vision-language model. Then two key tasks are addressed in this process by computing the cosine similarity between embeddings. First, label sanitization is performed to identify irrelevant, misspelled, or semantically weak labels, and surface the most semantically aligned label for each image by comparing image-label pairs using cosine similarity between image and label embeddings. Second, the method applies density-based clustering on text embeddings, followed by iterative cluster merging, to group semantically similar labels into unified label groups. The Factorynet dataset, which includes noisy labels from both human annotations and web-scraped sources, is employed to evaluate the effectiveness of the proposed framework. Experimental results demonstrate that the VLSR framework successfully identifies problematic labels and improves label consistency. This method enables a significant reduction in label vocabulary through clustering, which ultimately enhances the dataset's quality for training robust machine learning models in industrial applications with minimal human intervention.

cs.CV↗

arXiv2025

Quasisteady patterns in interfaces: Folding and Faceting

We present a systematic derivation of the gradient flows associated to a broad class of interfacial energies, emphasizing the relation between intrinsic and extrinsic variations of the interface. We show that the intrinsic variables formulation brings the gradient flow into alignment with the traditional analysis of quasi-steady dynamical systems defined on a stationary domain. Gradient flows are derived for model systems which exhibit quasi-steady pattern formation including coarsening among faceted interfaces and nonlocal interactions that model membrane self-adhesion and self-avoidance.

math.AP↗

arXiv2024

The Structure of Symbolic Powers of Matroids

We describe the structure of the symbolic powers $I^{(\ell)}$ of the Stanley-Reisner ideals, and cover ideals, $I$, of matroids. We (a) prove a structure theorem describing a minimal generating set for every $I^{(\ell)}$; (b) describe the (non--standard graded) symbolic Rees algebra $\mathcal{R}_s(I)$ of $I$ and show its minimal algebra generators have degree at most ht $I$; (c) provide an explicit, simple formula to compute the largest degree of a minimal algebra generator of $\mathcal{R}_s(I)$; (d) provide algebraic applications, including formulas for the symbolic defects of $I$, the initial degree of $I^{(\ell)}$, and the Waldschmidt constant of $I$; (e) provide a new algorithm allowing fast computations of very large symbolic powers of $I$. One of the by-products is a new characterization of matroids in terms of minimal generators of $I^{(\ell)}$ for some $\ell\geq 2$. In particular, it yields a new, simple characterization of matroids in terms of the minimal generators of $I^{(2)}$. This is the first characterization of matroids in terms of $I^{(2)}$, and it complements a celebrated theorem by Minh-Trung, Varbaro, and Terai-Trung which requires the investigation of homological properties of $I^{(\ell)}$ for some $\ell\geq 3$.

math.AC↗

arXiv2024

From Coupled Oscillators to Graph Neural Networks: Reducing Over-smoothing via a Kuramoto Model-based Approach

We propose the Kuramoto Graph Neural Network (KuramotoGNN), a novel class of continuous-depth graph neural networks (GNNs) that employs the Kuramoto model to mitigate the over-smoothing phenomenon, in which node features in GNNs become indistinguishable as the number of layers increases. The Kuramoto model captures the synchronization behavior of non-linear coupled oscillators. Under the view of coupled oscillators, we first show the connection between Kuramoto model and basic GNN and then over-smoothing phenomenon in GNNs can be interpreted as phase synchronization in Kuramoto model. The KuramotoGNN replaces this phase synchronization with frequency synchronization to prevent the node features from converging into each other while allowing the system to reach a stable synchronized state. We experimentally verify the advantages of the KuramotoGNN over the baseline GNNs and existing methods in reducing over-smoothing on various graph deep learning benchmark tasks.

cs.LG↗

arXiv2023

Solving the Right Problem is Key for Translational NLP: A Case Study in UMLS Vocabulary Insertion

As the immense opportunities enabled by large language models become more apparent, NLP systems will be increasingly expected to excel in real-world settings. However, in many instances, powerful models alone will not yield translational NLP solutions, especially if the formulated problem is not well aligned with the real-world task. In this work, we study the case of UMLS vocabulary insertion, an important real-world task in which hundreds of thousands of new terms, referred to as atoms, are added to the UMLS, one of the most comprehensive open-source biomedical knowledge bases. Previous work aimed to develop an automated NLP system to make this time-consuming, costly, and error-prone task more efficient. Nevertheless, practical progress in this direction has been difficult to achieve due to a problem formulation and evaluation gap between research output and the real-world task. In order to address this gap, we introduce a new formulation for UMLS vocabulary insertion which mirrors the real-world task, datasets which faithfully represent it and several strong baselines we developed through re-purposing existing solutions. Additionally, we propose an effective rule-enhanced biomedical language model which enables important new model behavior, outperforms all strong baselines and provides measurable qualitative improvements to editors who carry out the UVI task. We hope this case study provides insight into the considerable importance of problem formulation for the success of translational NLP solutions.

cs.CL↗

arXiv2023

p-Laplacian Transformer

$p$-Laplacian regularization, rooted in graph and image signal processing, introduces a parameter $p$ to control the regularization effect on these data. Smaller values of $p$ promote sparsity and interpretability, while larger values encourage smoother solutions. In this paper, we first show that the self-attention mechanism obtains the minimal Laplacian regularization ($p=2$) and encourages the smoothness in the architecture. However, the smoothness is not suitable for the heterophilic structure of self-attention in transformers where attention weights between tokens that are in close proximity and non-close ones are assigned indistinguishably. From that insight, we then propose a novel class of transformers, namely the $p$-Laplacian Transformer (p-LaT), which leverages $p$-Laplacian regularization framework to harness the heterophilic features within self-attention layers. In particular, low $p$ values will effectively assign higher attention weights to tokens that are in close proximity to the current token being processed. We empirically demonstrate the advantages of p-LaT over the baseline transformers on a wide range of benchmark datasets.

cs.LG↗

arXiv2023

Exponential relaxation to equilibrium for a kinetic Fokker-Planck-Alignment equation with force

In this note, we consider a kinetic Fokker-Planck-Alignment equation with Rayleigh-type friction and self-propulsion force which is derived from general environmental averaging models. We show the exponential relaxation in time toward equilibrium of the solutions provided certain spectral gap conditions are satisfied. The result is proved by using Desvillettes-Villani's method for collisional models to establish the global hypocoercivity.

math.AP↗

arXiv2023

Revisiting Over-smoothing and Over-squashing Using Ollivier-Ricci Curvature

Graph Neural Networks (GNNs) had been demonstrated to be inherently susceptible to the problems of over-smoothing and over-squashing. These issues prohibit the ability of GNNs to model complex graph interactions by limiting their effectiveness in taking into account distant information. Our study reveals the key connection between the local graph geometry and the occurrence of both of these issues, thereby providing a unified framework for studying them at a local scale using the Ollivier-Ricci curvature. Specifically, we demonstrate that over-smoothing is linked to positive graph curvature while over-squashing is linked to negative graph curvature. Based on our theory, we propose the Batch Ollivier-Ricci Flow, a novel rewiring algorithm capable of simultaneously addressing both over-smoothing and over-squashing.

cs.LG↗

arXiv2022

Continuous model of opinion dynamics with convictions

In this note we study a new kinetic model of opinion dynamics. The model incorporates two forces -- alignment of opinions under all-to-all communication driving the system to a consensus, and Rayleigh type friction force that drives each `player' to its fixed conviction value. The balance between these forces creates a non-trivial limiting outcome. We establish existence of a global mono-opinion state, whereby any initial distribution of opinions for each conviction value aggregates to the Dirac measure concentrated on a single opinion. We identify several cases where such a state is unique and depends continuously on the initial distribution of convictions. Several regularity properties of the limiting distribution of opinions are presented.

math.AP↗

arXiv2022

VidConv: A modernized 2D ConvNet for Efficient Video Recognition

Since being introduced in 2020, Vision Transformers (ViT) has been steadily breaking the record for many vision tasks and are often described as ``all-you-need" to replace ConvNet. Despite that, ViTs are generally computational, memory-consuming, and unfriendly for embedded devices. In addition, recent research shows that standard ConvNet if redesigned and trained appropriately can compete favorably with ViT in terms of accuracy and scalability. In this paper, we adopt the modernized structure of ConvNet to design a new backbone for action recognition. Particularly, our main target is to serve for industrial product deployment, such as FPGA boards in which only standard operations are supported. Therefore, our network simply consists of 2D convolutions, without using any 3D convolution, long-range attention plugin, or Transformer blocks. While being trained with much fewer epochs (5x-10x), our backbone surpasses the methods using (2+1)D and 3D convolution, and achieve comparable results with ViT on two benchmark datasets.

cs.CV↗

arXiv2022

Generic Generalized Diagonal Matrices

Generalized diagonal matrices are matrices that have two ladders of entries that are zero in the upper right and bottom left corners. The minors of generic generalized diagonal matrices have square-free initial ideals. We give a description of the facets of their Stanley-Reisner complex. With this description, we characterize the configuration of ladders that yield Cohen-Macaulay ideals. In the special case where both ladders are triangles, we show that the corresponding complex is vertex decomposable. Also in this case, we compute the height and multiplicity of the ideals.

math.AC↗

arXiv2022

UVA Resources for the Biomedical Vocabulary Alignment at Scale in the UMLS Metathesaurus

The construction and maintenance process of the UMLS (Unified Medical Language System) Metathesaurus is time-consuming, costly, and error-prone as it relies on (1) the lexical and semantic processing for suggesting synonymous terms, and (2) the expertise of UMLS editors for curating the suggestions. For improving the UMLS Metathesaurus construction process, our research group has defined a new task called UVA (UMLS Vocabulary Alignment) and generated a dataset for evaluating the task. Our group has also developed different baselines for this task using logical rules (RBA), and neural networks (LexLM and ConLM). In this paper, we present a set of reusable and reproducible resources including (1) a dataset generator, (2) three datasets generated by using the generator, and (3) three baseline approaches. We describe the UVA dataset generator and its implementation generalized for any given UMLS release. We demonstrate the use of the dataset generator by generating datasets corresponding to three UMLS releases, 2020AA, 2021AA, and 2021AB. We provide three UVA baselines using the three existing approaches (LexLM, ConLM, and RBA). The code, the datasets, and the experiments are publicly available, reusable, and reproducible with any UMLS release (a no-cost license agreement is required for downloading the UMLS).

cs.AI↗

arXiv2022

UBERT: A Novel Language Model for Synonymy Prediction at Scale in the UMLS Metathesaurus

The UMLS Metathesaurus integrates more than 200 biomedical source vocabularies. During the Metathesaurus construction process, synonymous terms are clustered into concepts by human editors, assisted by lexical similarity algorithms. This process is error-prone and time-consuming. Recently, a deep learning model (LexLM) has been developed for the UMLS Vocabulary Alignment (UVA) task. This work introduces UBERT, a BERT-based language model, pretrained on UMLS terms via a supervised Synonymy Prediction (SP) task replacing the original Next Sentence Prediction (NSP) task. The effectiveness of UBERT for UMLS Metathesaurus construction process is evaluated using the UMLS Vocabulary Alignment (UVA) task. We show that UBERT outperforms the LexLM, as well as biomedical BERT-based models. Key to the performance of UBERT are the synonymy prediction task specifically developed for UBERT, the tight alignment of training data to the UVA task, and the similarity of the models used for pretrained UBERT.

cs.CL↗

arXiv2021

Propagation of Chaos for the Cucker-Smale Systems under heavy tail communication

In this work we study propagation of chaos for solutions of the Liouville equation for the classical discrete Cucker-Smale system. Assuming that the communication kernel satisfies the heavy tail condition -- known to be necessary to induce exponential alignment -- we obtain a linear in time convergence rate of the $k$-th marginals $f^{(k)}$ to the product of $k$ solutions of the corresponding Vlasov-Alignment equation, $f^{\otimes k}$. Specifically, the following estimate holds in terms of Wasserstein-2 metric \begin{equation}\label{e:abst} \mathcal{W}_2(f^{(k)}_t,f^{\otimes k}_t) \leq C k^{1/2} \min\left\{ 1, \frac{t}{\sqrt{N}} \right\}. \end{equation} For systems with the Rayleigh-type friction and self-propulsion force, we obtain a similar result for sectorial solutions. Such solutions are known to align exponentially fast via the method of Grassmannian reduction, \cite{LRS-friction}. We recast the method in the kinetic settings and show that the bound \eqref{e:abst} persists but with the quadratic dependence on time. In both the forceless and forced cases, the result represents an improvement over the exponential bounds established earlier in the work of Natalini and Paul, \cite{NP2021-orig}, although those bounds hold for general kernels. The main message of our work is that flocking dynamics improves the rate considerably.

math.AP↗

arXiv2021

Evaluating Biomedical BERT Models for Vocabulary Alignment at Scale in the UMLS Metathesaurus

The current UMLS (Unified Medical Language System) Metathesaurus construction process for integrating over 200 biomedical source vocabularies is expensive and error-prone as it relies on the lexical algorithms and human editors for deciding if the two biomedical terms are synonymous. Recent advances in Natural Language Processing such as Transformer models like BERT and its biomedical variants with contextualized word embeddings have achieved state-of-the-art (SOTA) performance on downstream tasks. We aim to validate if these approaches using the BERT models can actually outperform the existing approaches for predicting synonymy in the UMLS Metathesaurus. In the existing Siamese Networks with LSTM and BioWordVec embeddings, we replace the BioWordVec embeddings with the biomedical BERT embeddings extracted from each BERT model using different ways of extraction. In the Transformer architecture, we evaluate the use of the different biomedical BERT models that have been pre-trained using different datasets and tasks. Given the SOTA performance of these BERT models for other downstream tasks, our experiments yield surprisingly interesting results: (1) in both model architectures, the approaches employing these biomedical BERT-based models do not outperform the existing approaches using Siamese Network with BioWordVec embeddings for the UMLS synonymy prediction task, (2) the original BioBERT large model that has not been pre-trained with the UMLS outperforms the SapBERT models that have been pre-trained with the UMLS, and (3) using the Siamese Networks yields better performance for synonymy prediction when compared to using the biomedical BERT models.

cs.CL↗

arXiv2020

Lech's Inequality for the Buchsbaum-Rim Multiplicity and Mixed Multiplicity

We generalize an improved Lech bound, due to Huneke, Smirnov, and Validashti, for the Buchsbaum-Rim multiplicity and mixed multiplicity. We reduce the problem to the graded case and then to the polynomial ring case. There we use complete reductions, studied by Rees, to prove sharper bounds for the mixed multiplicity in low dimensions before proving the general case.

math.AC↗

arXiv2019

Alternative Blockmodelling

Many approaches have been proposed to discover clusters within networks. Community finding field encompasses approaches which try to discover clusters where nodes are tightly related within them but loosely related with nodes of other clusters. However, a community network configuration is not the only possible latent structure in a graph. Core-periphery and hierarchical network configurations are valid structures to discover in a relational dataset. On the other hand, a network is not completely explained by only knowing the membership of each node. A high level view of the inter-cluster relationships is needed. Blockmodelling techniques deal with these two issues. Firstly, blockmodelling allows finding any network configuration besides to the well-known community structure. Secondly, blockmodelling is a summary representation of a network which regards not only membership of nodes but also relations between clusters. Finally, a unique summary representation of a network is unlikely. Networks might hide more than one blockmodel. Therefore, our proposed problem aims to discover a secondary blockmodel representation of a network that is of good quality and dissimilar with respect to a given blockmodel. Our methodology is presented through two approaches, (a) inclusion of cannot-link constraints and (b) dissimilarity between image matrices. Both approaches are based on non-negative matrix factorisation NMF which fits the blockmodelling representation. The evaluation of these two approaches regards quality and dissimilarity of the discovered alternative blockmodel as these are the requirements of the problem.

cs.SI↗