arXiv subjects

Trung Thanh Nguyen

Publications and source records attributed to Trung Thanh Nguyen.

35 records · Page 2Linked to original sources

arXiv2024

CT to PET Translation: A Large-scale Dataset and Domain-Knowledge-Guided Diffusion Approach

Positron Emission Tomography (PET) and Computed Tomography (CT) are essential for diagnosing, staging, and monitoring various diseases, particularly cancer. Despite their importance, the use of PET/CT systems is limited by the necessity for radioactive materials, the scarcity of PET scanners, and the high cost associated with PET imaging. In contrast, CT scanners are more widely available and significantly less expensive. In response to these challenges, our study addresses the issue of generating PET images from CT images, aiming to reduce both the medical examination cost and the associated health risks for patients. Our contributions are twofold: First, we introduce a conditional diffusion model named CPDM, which, to our knowledge, is one of the initial attempts to employ a diffusion model for translating from CT to PET images. Second, we provide the largest CT-PET dataset to date, comprising 2,028,628 paired CT-PET images, which facilitates the training and evaluation of CT-to-PET translation models. For the CPDM model, we incorporate domain knowledge to develop two conditional maps: the Attention map and the Attenuation map. The former helps the diffusion process focus on areas of interest, while the latter improves PET data correction and ensures accurate diagnostic information. Experimental evaluations across various benchmarks demonstrate that CPDM surpasses existing methods in generating high-quality PET images in terms of multiple metrics. The source code and data samples are available at https://github.com/thanhhff/CPDM.

eess.IV↗

arXiv2024

Action Selection Learning for Multi-label Multi-view Action Recognition

Multi-label multi-view action recognition aims to recognize multiple concurrent or sequential actions from untrimmed videos captured by multiple cameras. Existing work has focused on multi-view action recognition in a narrow area with strong labels available, where the onset and offset of each action are labeled at the frame-level. This study focuses on real-world scenarios where cameras are distributed to capture a wide-range area with only weak labels available at the video-level. We propose the method named Multi-view Action Selection Learning (MultiASL), which leverages action selection learning to enhance view fusion by selecting the most useful information from different viewpoints. The proposed method includes a Multi-view Spatial-Temporal Transformer video encoder to extract spatial and temporal features from multi-viewpoint videos. Action Selection Learning is employed at the frame-level, using pseudo ground-truth obtained from weak labels at the video-level, to identify the most relevant frames for action recognition. Experiments in a real-world office environment using the MM-Office dataset demonstrate the superior performance of the proposed method compared to existing methods. The source code is available at https://github.com/thanhhff/MultiASL/.

cs.CV↗

arXiv2024

FedCert: Federated Accuracy Certification

Federated Learning (FL) has emerged as a powerful paradigm for training machine learning models in a decentralized manner, preserving data privacy by keeping local data on clients. However, evaluating the robustness of these models against data perturbations on clients remains a significant challenge. Previous studies have assessed the effectiveness of models in centralized training based on certified accuracy, which guarantees that a certain percentage of the model's predictions will remain correct even if the input data is perturbed. However, the challenge of extending these evaluations to FL remains unresolved due to the unknown client's local data. To tackle this challenge, this study proposed a method named FedCert to take the first step toward evaluating the robustness of FL systems. The proposed method is designed to approximate the certified accuracy of a global model based on the certified accuracy and class distribution of each client. Additionally, considering the Non-Independent and Identically Distributed (Non-IID) nature of data in real-world scenarios, we introduce the client grouping algorithm to ensure reliable certified accuracy during the aggregation step of the approximation algorithm. Through theoretical analysis, we demonstrate the effectiveness of FedCert in assessing the robustness and reliability of FL systems. Moreover, experimental results on the CIFAR-10 and CIFAR-100 datasets under various scenarios show that FedCert consistently reduces the estimation error compared to baseline methods. This study offers a solution for evaluating the robustness of FL systems and lays the groundwork for future research to enhance the dependability of decentralized learning. The source code is available at https://github.com/thanhhff/FedCert/.

cs.LG↗

arXiv2024

A Tangible Multi-Display Toolkit to Support the Collaborative Design Exploration of AV-Pedestrian Interfaces

The advent of cyber-physical systems, such as robots and autonomous vehicles (AVs), brings new opportunities and challenges for the domain of interaction design. Though there is consensus about the value of human-centred development, there is a lack of documented tailored methods and tools for involving multiple stakeholders in design exploration processes. In this paper we present a novel approach using a tangible multi-display toolkit. Orchestrating computer-generated imagery across multiple displays, the toolkit enables multiple viewing angles and perspectives to be captured simultaneously (e.g. top-view, first-person pedestrian view). Participants are able to directly interact with the simulated environment through tangible objects. At the same time, the objects physically simulate the interface's behaviour (e.g. through an integrated LED display). We evaluated the toolkit in design sessions with experts to collect feedback and input on the design of an AV-pedestrian interface. The paper reports on how the combination of tangible objects and multiple displays supports collaborative design explorations.

cs.HC↗

arXiv2024

Q-learning-based Opportunistic Communication for Real-time Mobile Air Quality Monitoring Systems

We focus on real-time air quality monitoring systems that rely on devices installed on automobiles in this research. We investigate an opportunistic communication model in which devices can send the measured data directly to the air quality server through a 4G communication channel or via Wi-Fi to adjacent devices or the so-called Road Side Units deployed along the road. We aim to reduce 4G costs while assuring data latency, where the data latency is defined as the amount of time it takes for data to reach the server. We propose an offloading scheme that leverages Q-learning to accomplish the purpose. The experiment results show that our offloading method significantly cuts down around 40-50% of the 4G communication cost while keeping the latency of 99.5% packets smaller than the required threshold.

cs.NI↗

arXiv2024

Fuzzy Q-Learning-Based Opportunistic Communication for MEC-Enhanced Vehicular Crowdsensing

This study focuses on MEC-enhanced, vehicle-based crowdsensing systems that rely on devices installed on automobiles. We investigate an opportunistic communication paradigm in which devices can transmit measured data directly to a crowdsensing server over a 4G communication channel or to nearby devices or so-called Road Side Units positioned along the road via Wi-Fi. We tackle a new problem that is how to reduce the cost of 4G while preserving the latency. We propose an offloading strategy that combines a reinforcement learning technique known as Q-learning with Fuzzy logic to accomplish the purpose. Q-learning assists devices in learning to decide the communication channel. Meanwhile, Fuzzy logic is used to optimize the reward function in Q-learning. The experiment results show that our offloading method significantly cuts down around 30-40% of the 4G communication cost while keeping the latency of 99% packets below the required threshold.

cs.NI↗

arXiv2024

One-Stage Open-Vocabulary Temporal Action Detection Leveraging Temporal Multi-scale and Action Label Features

Open-vocabulary Temporal Action Detection (Open-vocab TAD) is an advanced video analysis approach that expands Closed-vocabulary Temporal Action Detection (Closed-vocab TAD) capabilities. Closed-vocab TAD is typically confined to localizing and classifying actions based on a predefined set of categories. In contrast, Open-vocab TAD goes further and is not limited to these predefined categories. This is particularly useful in real-world scenarios where the variety of actions in videos can be vast and not always predictable. The prevalent methods in Open-vocab TAD typically employ a 2-stage approach, which involves generating action proposals and then identifying those actions. However, errors made during the first stage can adversely affect the subsequent action identification accuracy. Additionally, existing studies face challenges in handling actions of different durations owing to the use of fixed temporal processing methods. Therefore, we propose a 1-stage approach consisting of two primary modules: Multi-scale Video Analysis (MVA) and Video-Text Alignment (VTA). The MVA module captures actions at varying temporal resolutions, overcoming the challenge of detecting actions with diverse durations. The VTA module leverages the synergy between visual and textual modalities to precisely align video segments with corresponding action labels, a critical step for accurate action identification in Open-vocab scenarios. Evaluations on widely recognized datasets THUMOS14 and ActivityNet-1.3, showed that the proposed method achieved superior results compared to the other methods in both Open-vocab and Closed-vocab settings. This serves as a strong demonstration of the effectiveness of the proposed method in the TAD task.

cs.CV↗

arXiv2024

Designing for Projection-based Communication between Autonomous Vehicles and Pedestrians

Recent studies have investigated new approaches for communicating an autonomous vehicle's (AV) intent and awareness to pedestrians. This paper adds to this body of work by presenting the design and evaluation of in-situ projections on the road. Our design combines common traffic light patterns with aesthetic visual elements. We describe the iterative design process and the prototyping methods used in each stage. The final design concept was represented as a virtual reality simulation and evaluated with 18 participants in four different street crossing scenarios, which included three scenarios that simulated various degrees of system errors. We found that different design elements were able to support participants' confidence in their decision even when the AV failed to correctly detect their presence. We also identified elements in our design that needed to be more clearly communicated. Based on these findings, the paper presents a series of design recommendations for projection-based communication between AVs and pedestrians.

cs.HC↗

arXiv2023

Improved Approximation Guarantees for Power Scheduling Problems With Sum-of-Squares Constraints

We study a class of combinatorial scheduling problems characterized by a particular type of constraint often associated with electrical power or gas energy. This constraint appears in several practical applications and is expressed as a sum of squares of linear functions. Its nonlinear nature adds complexity to the scheduling problem, rendering it notably challenging, even in the case of a linear objective. In fact, exact polynomial time algorithms are unlikely to exist, and thus, prior works have focused on designing approximation algorithms with polynomial running time and provable guarantees on the solution quality. In an effort to advance this line of research, we present novel approximation algorithms yielding significant improvements over the existing state-of-the-art results for these problems.

cs.DS↗

arXiv2023

Joint Rate Allocation and Power Control for RSMA-Based Communication and Radar Coexistence Systems

We consider a rate-splitting multiple access (RSMA)-based communication and radar coexistence (CRC) system. The proposed system allows an RSMA-based communication system to share spectrum with multiple radars. Furthermore, RSMA enables flexible and powerful interference management by splitting messages into common parts and private parts to partially decode interference and partially treat interference as noise. The RSMA-based CRC system thus significantly improves spectral efficiency, energy efficiency and quality of service (QoS) of communication users (CUs). However, the RSMA-based CRC system raises new challenges. Due to the spectrum sharing, the communication network and the radars cause interference to each other, which reduces the signal-to-interference-plus-noise ratio (SINR) of the radars as well as the data rate of the CUs in the communication network. Therefore, a major problem is to maximize the sum rate of the CUs while guaranteeing their QoS requirements of data transmissions and the SINR requirements of multiple radars. To achieve these objectives, we formulate a problem that optimizes i) the common rate allocation to the CUs, transmit power of common messages and transmit power of private messages of the CUs, and ii) transmit power of the radars. The problem is non-convex with multiple decision parameters, which is challenging to be solved. We propose two algorithms. The first sequential quadratic programming (SQP) can quickly return a local optimal solution, and has been known to be the state-of-the-art in nonlinear programming methods. The second is an additive approximation scheme (AAS) which solves the problem globally in a reasonable amount of time, based on the technique of applying exhaustive enumeration to a modified instance. The simulation results show the improvement of the AAS compared with the SQP in terms of sum rate.

cs.IT↗

arXiv2022

A Novel Approach for Pill-Prescription Matching with GNN Assistance and Contrastive Learning

Medication mistaking is one of the risks that can result in unpredictable consequences for patients. To mitigate this risk, we develop an automatic system that correctly identifies pill-prescription from mobile images. Specifically, we define a so-called pill-prescription matching task, which attempts to match the images of the pills taken with the pills' names in the prescription. We then propose PIMA, a novel approach using Graph Neural Network (GNN) and contrastive learning to address the targeted problem. In particular, GNN is used to learn the spatial correlation between the text boxes in the prescription and thereby highlight the text boxes carrying the pill names. In addition, contrastive learning is employed to facilitate the modeling of cross-modal similarity between textual representations of pill names and visual representations of pill images. We conducted extensive experiments and demonstrated that PIMA outperforms baseline models on a real-world dataset of pill and prescription images that we constructed. Specifically, PIMA improves the accuracy from 19.09% to 46.95% compared to other baselines. We believe our work can open up new opportunities to build new clinical applications and improve medication safety and patient care.

cs.CV↗

arXiv2022

FedDRL: Deep Reinforcement Learning-based Adaptive Aggregation for Non-IID Data in Federated Learning

The uneven distribution of local data across different edge devices (clients) results in slow model training and accuracy reduction in federated learning. Naive federated learning (FL) strategy and most alternative solutions attempted to achieve more fairness by weighted aggregating deep learning models across clients. This work introduces a novel non-IID type encountered in real-world datasets, namely cluster-skew, in which groups of clients have local data with similar distributions, causing the global model to converge to an over-fitted solution. To deal with non-IID data, particularly the cluster-skewed data, we propose FedDRL, a novel FL model that employs deep reinforcement learning to adaptively determine each client's impact factor (which will be used as the weights in the aggregation process). Extensive experiments on a suite of federated datasets confirm that the proposed FedDRL improves favorably against FedAvg and FedProx methods, e.g., up to 4.05% and 2.17% on average for the CIFAR-100 dataset, respectively.

cs.LG↗

arXiv2021

Access Management in Joint Sensing and Communication Systems: Efficiency versus Fairness

In this paper, we consider a distributed joint sensing and communication (DJSC) system in which multiple radar sensors are deployed. Each sensor is equipped with a sensing function and a communication function, and thus it is a JSC node. The JSC nodes are able to perform sensing their surrounding environments, e.g., weather conditions or available spectrum. Furthermore, they can cooperatively detect and track a common target. The information, i.e., of the environment and target, collected by the JSC nodes is transmitted to a base station (BS) for further processing. As such, different aspects of the target to be viewed simultaneously, which significantly improves the performance of the target detection and tracking. However, both the sensing function and communication function require a certain amount of bandwidth for their operations, and deploying multiple JSC nodes may consume a large amount of bandwidth. Therefore, we investigate the bandwidth allocation problem for the DJSC system. In particular, we aim to optimize the bandwidth allocation to the sensing function and the communication function of the JSC nodes. To improve the allocation efficiency while benefiting the spatial diversity advantage of the DJSC systems, the objective is to maximize the sum of sensing performances, i.e., estimation rates, communication performances, i.e., communication data rates, and fairnesses of all the users. The optimization problem is non-convex and difficult to be solved. For this, we propose a fully polynomial time approximation algorithm, and we prove that the approximation algorithm can guarantee a near-optimal solution with an accuracy bound of $ε$. Furthermore, we propose to use a heuristic algorithm with lower complexity. The simulation results show that both the proposed algorithms are able to achieve the solutions close to the optimum in a computationally efficient fashion.

cs.NI↗

arXiv2021

Exact Algorithms for Scheduling Problems on Parallel Identical Machines with Conflict Jobs

Machine scheduling problems involving conflict jobs can be seen as a constrained version of the classical scheduling problem, in which some jobs are conflict in the sense that they cannot be proceeded simultaneously on different machines. This conflict constraint naturally arises in several practical applications and has recently received considerable attentions in the research community. In fact, the problem is typically NP-hard (even for approximation) and most of algorithmic results achieved so far have heavily relied on special structures of the underlying graph used to model the conflict-job relation. Our focus is on three objective functions: minimizing the makespan, minimizing the weighted summation of the jobs' completion time, and maximizing the total weights of completed jobs; the first two of which have been intensively studied in the literature. For each objective function considered, we present several mixed integer linear programming models and a constraint programming model, from which we can solve the problems to optimality using dedicated solvers. Binary search-based algorithms are also proposed to solve the makespan problem. The results of numerical experiments performed on randomly generated data sets with up to 32 jobs and 6 machines are reported and analysed to verify the performance of the proposed methods.

cs.DS↗

arXiv2016

Dynamic Pricing in Smart Grids under Thresholding Policies: Algorithms and Heuristics

Minimizing the peak power consumption and matching demand to supply, under fixed threshold polices, are two key requirements for the success of the future electricity market. In this work, we consider dynamic pricing methods to minimize the peak load and match demand to supply in the smart grid. As these optimization problems are computationally hard to solve in general, we propose generic heuristics for approximating their solutions. Further, we provide theoretical analysis of uniform pricing in peak-demand minimization. Moreover, we propose optimal-pricing algorithms for scenarios in which the time-period in which tasks must be executed is relatively small. Finally, we conduct several experiments to evaluate the various algorithms on real data.

math.OC↗

arXiv2014

Approximation Schemes for Binary Quadratic Programming Problems with Low cp-Rank Decompositions

Binary quadratic programming problems have attracted much attention in the last few decades due to their potential applications. This type of problems are NP-hard in general, and still considered a challenge in the design of efficient approximation algorithms for their solutions. The purpose of this paper is to investigate the approximability for a class of such problems where the constraint matrices are {\it completely positive} and have low {\it cp-rank}. In the first part of the paper, we show that a completely positive rational factorization of such matrices can be computed in polynomial time, within any desired accuracy. We next consider binary quadratic programming problems of the following form: Given matrices $Q_1,...,Q_n\in\mathbb{R}_+^{n\times n}$, and a system of $m$ constrains $x^TQ_ix\le C_i^2$ ($x^TQ_ix\ge C_i^2$), $i=1,...,m$, we seek to find a vector $x^*\in \{0,1\}^n$ that maximizes (minimizes) a given function $f$. This class of problems generalizes many fundamental problems in discrete optimization such as packing and covering integer programs/knapsack problems, quadratic knapsack problems, submodular maximization, etc. We consider the case when $m$ and the cp-ranks of the matrices $Q_i$ are bounded by a constant. Our approximation results for the maximization problem are as follows. For the case when the objective function is nonnegative submodular, we give an $(1/4-ε)$-approximation algorithm, for any $ε>0$; when the function $f$ is linear, we present a PTAS. We next extend our PTAS result to a wider class of non-linear objective functions including quadratic functions, multiplicative functions, and sum-of-ratio functions. The minimization problem seems to be much harder due to the fact that the relaxation is {\it not} convex. For this case, we give a QPTAS for $m=1$.

cs.DS↗

arXiv2014

Global optimization using Lévy flights

This paper studies a class of enhanced diffusion processes in which random walkers perform Lévy flights and apply it for global optimization. Lévy flights offer controlled balance between exploitation and exploration. We develop four optimization algorithms based on such properties. We compare new algorithms with the well-known Simulated Annealing on hard test functions and the results are very promising.

cs.NE↗