Artificial Intelligence and Robotics Commons^™

Research On Inpainting Algorithm Of Digital Murals Based On Enhanced Structural Information, Ziying Zhang, Hua Zhou

Journal of System Simulation

Abstract: According to the fact that the murals of Fahai Temple in Beijing are missing in blocks and the missing area is structure information, a structure-enhancing digital image restoration algorithm is proposed to solve the problem of insufficient consideration of image structure information in Criminisi algorithm. When calculating the priority function of the filling block, the curvature calculation of the linear convolution is integrated into the data item, and the weight of the structure information is increased to achieve the goal of repairing the structure information-rich region in priority; the regional covariance method is introduced in the similarity calculation of …

Simulation Research On Covid-19 Transmission And Control Measures Based On SeiIRd Model, Jing Wang, Ying Dong

Journal of System Simulation

Abstract: With the spread of the novel coronavirus pneumonia around the world, the data and transmission mechanism are analyzed. The SEI_iRD model is constructed based on the existing SEIRD model, and the infected population is divided into asymptomatic infections, mild infections, severe infections and critical infections. The impact of the transmission rate of different infected people on the development of the epidemic was analyzed. Simulation experiments were carried out on the basis of fitting real data, and it was found that the main infected populations that affected the discovery of the epidemic were asymptomatic and mildly infected. On …

Interactive Construction Of Scientific Workflow Based On Process Mining, Jun Liu, Yang Gao, Tao Xu, Qing Zhao, Guihua Shan, Xuebin Chi

Journal of System Simulation

Abstract: When dealing with large-scale or complex workflows, the construction efficiency of traditional interactive workflow construction methods is very low. To solve this problem, a workflow construction method based on process mining is proposed. Heuristic methods are used to collect process fragments. The specially designed relation description language is used to record the process description of different levels and aspects in the workflow as text. The text is translated to generate process relational data, which will be output to the process discovery algorithm to generate a sound workflow network. An interactive workflow construction software has been developed and tested in …

A Quantization Training Algorithm Of Adaptive Learning Quantization Scale Fators, Hui Nie, Kangshun Li, Yang Su

Journal of System Simulation

Abstract: Deep neural network model is difficult to effectively deploy in embedded terminals due to its excessive number of components, andone of the solutions is model miniaturization (such as model quantization, knowledge distillation, etc.). To address this problem, a quantization training algorithm (referred to as LSQ-BN algorithm) based on adaptive learning of quantizationscale factors with BN folding is proposed.A single CNN (convolutional neural) is usedtoconstruct BN folding and achieve BN and CNN fusion. During the process of quantitative training,the quantization scale factors are set as model parameters. An adaptive quantizationscale factor initialization scheme is proposed to solve the problem …

Joint Shift Scheduling Method For Call Center With Mechanism Of Delay Information, Miao Yu, Manru Li, Yu Zhao

Journal of System Simulation

Abstract: A joint shift scheduling method is studied for call center with delay information. According to the queue model of call center with delay information, the influence rule of the customer's patience and abandonment behavior is addressed, and a mechanism of delay information is proposed to estimate the waiting time of customers. Considering the influence of non-stationary arrival and other factors, the scheduling model of the call centers is established by the discrete Event-Scheduling approach. Based on the proposed evaluation method of delay information, the joint shift scheduling method by simulation optimization is designed to solve the scheduling problem …

Modeling And Simulation Of Ultra Supercritical Unit Using A Composite Weighted Human Learning Network, Chuanliang Cheng, Chen Peng, Deliang Zeng, Tengfei Zhang

Journal of System Simulation

Abstract: Intermediate point temperature is an important parameter in ultra supercritical (USC) unit. However, due to strong nonlinearity, it is difficult to determine the form and coefficients of the corresponding model by using traditional methods. In order to get a better control effect, a novel composite weighted human learning optimization network (CWHLON) is proposed to tackle the above-mentioned problems. Though the real-time dynamic linear model, the characteristics of the object are accurately simulated. In the simulation experiment, CWHLON is compared with the traditional recursive least squares and other three meta heuristic methods. The data show that the proposed method improves …

Research On The Number Of Passengers On The Platform Of Rail Transit Station Considering Congestion Propagation, Wei Chen, Zongping Li, Can Liu, Yanni Ju

Journal of System Simulation

Abstract: It is the basis of improving the safety guarantee ability of urban rail transit system to study and master the change law of the number of passengers in the urban rail transit station under the condition of Congestion Propagation. From the point of view of multi subsystem of passenger, station and train, combined with the multi-attribute characteristics of passenger flow, platform and train, the calculation model of the number of passengers in urban rail transit station is established based on system dynamics. A multi group sensitivity simulation experiment is designed to analyze the influence factors of the number of …

Multi-Stage Multi-Agv Path Planning With Walk Under Shelves For Robotic Mobile Fulfillment Systems, Teng Li, Peipei Ding, Jinfang Liu

Journal of System Simulation

Abstract: Aiming at the problem of increasing travel time due to turning and obstacle avoidance in robotic mobile fulfillment systems(RMFS) with large-scale multi-AGV path planning, a path planning model with the shortest task completion time is established. A path planning model considering no-load AGV that can pass through the shelf is proposed, and the model is solved by an improving A* algorithm. The AGV operation stage is divided, an turning penalty value is introduced into the A* algorithm to reduce the turning times, and the obstacle avoidance priority with the obstacle avoidance waiting time is set. The simulation results show …

Design And Simulation Of Ts Fuzzy Based Cooperative Control Of Missile Formation, Yexin Zhang, Yu Cheng, Hongyan Yan, Xuwei Fan, Xu Zhang, Yi Tian

Journal of System Simulation

Abstract: Aiming at the requirement of cooperative operation of multi-missile formation, a cooperative control algorithm of multi-missile formation based on Takagi-Sugeno(TS) fuzzy control theory is proposed.The flight speed, trajectory angle and trajectory deflection angle of the missile are taken as parameters in the leader-follower mode missile formation flying system.The local asymptotically stable controller is designed by using the systemlocal linearization of multiple groups of equilibrium pointsduring the whole flight process.Through the expert experience method,the membership function and fuzzy rules for the system are designedwith TS fuzzy theory, and the whole multi-missile cooperative control system is completed and the stability …

Green, Quantized Federated Learning Over Wireless Networks: An Energy-Efficient Design, Minsu Kim, Walid Saad, Mohammad Mozaffari, Mérouane Debbah

Machine Learning Faculty Publications

The practical deployment of federated learning (FL) over wireless networks requires balancing energy efficiency and convergence time due to the limited available resources of devices. Prior art on FL often trains deep neural networks (DNNs) to achieve high accuracy and fast convergence using 32 bits of precision level. However, such scenarios will be impractical for resource-constrained devices since DNNs typically have high computational complexity and memory requirements. Thus, there is a need to reduce the precision level in DNNs to reduce the energy expenditure. In this paper, a green-quantized FL framework, which represents data with a finite precision level in …

Adversarial Pixel Restoration As A Pretext Task For Transferable Perturbations, Hashmat Shadab Malik, Shahina K. Kunhimon, Muzammal Nasser, Salman Khan, Fahad Shahbaz Khan

Computer Vision Faculty Publications

Transferable adversarial attacks optimize adversaries from a pretrained surrogate model and known label space to fool the unknown black-box models. Therefore, these attacks are restricted by the availability of an effective surrogate model. In this work, we relax this assumption and propose Adversarial Pixel Restoration as a self-supervised alternative to train an effective surrogate model from scratch under the condition of no labels and few data samples. Our training approach is based on a min-max objective which reduces overfitting via an adversarial objective and thus optimizes for a more generalizable surrogate model. Our proposed attack is complimentary to our adversarial …

Robustar: Interactive Toolbox Supporting Precise Data Annotation For Robust Vision Learning, Chonghan Chen, Haohan Wang, Leyang Hu, Yuhao Zhang, Shuguang Lyu, Jingcheng Wu, Xinnuo Li, Linjing Sun, Eric Xing

Machine Learning Faculty Publications

We introduce the initial release of our software Robustar, which aims to improve the robustness of vision classification machine learning models through a data-driven perspective. Building upon the recent understanding that the lack of machine learning model’s robustness is the tendency of the model’s learning of spurious features, we aim to solve this problem from its root at the data perspective by removing the spurious features from the data before training. In particular, we introduce a software that helps the users to better prepare the data for training image classification models by allowing the users to annotate the spurious features …

Machine Learning Approach For Classifying Power Outage In Secondary Electric Distribution Network, Stephan Mgaya, Hellen Maziku

Tanzania Journal of Engineering and Technology (TJET)

Power outage is the problem that hinders social and economic development especially for developing countries like Tanzania. Frequent power outages damage electric equipment, and negatively affect the industrial production process. Power outages cannot be completely eradicated due to uncontrolled cause like natural calamities but technical challenges can be managed and hence reducing power outages. The existing manual methods used to locate power outage like customer calls is inefficient and time consuming. On the other hand, modern method like the Advanced Metering Infrastructure (AMI) still faces a challenge in effectively classifying power line outage due to the nature of imbalanced datasets. …

Big Data Analytics Framework For Effective Higher Education Institutions, George Matto

Tanzania Journal of Engineering and Technology (TJET)

There has been an increased dependency on Information and Communication Technologies (ICTs) in undertaking various activities in Higher Education Institutions (HEIs) ecosystems. Because of that, huge volumes of data have increasingly been generated. There have been, for instance, considerable amounts of data generated through electronic platforms involved in students’ admission and registration process, students’ academic records management, teaching and learning data, curriculum related data, and several other administrative data. Analysis of data generated from these platforms stands to give students, lecturers, HEIs Management, policy makers and implementers, and other stakeholders useful insights that would help in improving HEIs’ effectiveness. Unfortunately, …

Toward A Standard Formal Semantic Representation Of The Model Card Report, Muhammad Tuan Amith, Licong Cui, Degui Zhi, Kirk Roberts, Xiaoqian Jiang, Fang Li, Evan Yu, Cui Tao

Faculty, Staff and Student Publications

BACKGROUND: Model card reports aim to provide informative and transparent description of machine learning models to stakeholders. This report document is of interest to the National Institutes of Health's Bridge2AI initiative to address the FAIR challenges with artificial intelligence-based machine learning models for biomedical research. We present our early undertaking in developing an ontology for capturing the conceptual-level information embedded in model card reports.

RESULTS: Sourcing from existing ontologies and developing the core framework, we generated the Model Card Report Ontology. Our development efforts yielded an OWL2-based artifact that represents and formalizes model card report information. The current release of …

Control And Planning For Mobile Manipulators Used In Large Scale Manufacturing Processes, Joshua T. Nguyen

LSU Master's Theses

Sanding operations in industry is one of the few manufacturing tasks that has yet to achieve automation. Sanding tasks require skilled operators that have developed a sense of when a work piece is sufficiently sanded. In order to achieve automation in sanding with robotic systems, this developed sense, or intelligence, that human operators have needs to be understood and implemented in order to achieve, at the minimum, the same quality of work. The system will also need to have the equivalent reach of a human operator and not be constrained to a single, small workspace. This thesis developed solutions for …

Bridging The Gap Between Object And Image-Level Representations For Open-Vocabulary Detection, Hanoona Rasheed, Muhammad Maaz, Muhammad Uzair Khattak, Salman Khan, Fahad Shahbaz Khan

Computer Vision Faculty Publications

Existing open-vocabulary object detectors typically enlarge their vocabulary sizes by leveraging different forms of weak supervision. This helps generalize to novel objects at inference. Two popular forms of weak-supervision used in open-vocabulary detection (OVD) include pretrained CLIP model and image-level supervision. We note that both these modes of supervision are not optimally aligned for the detection task: CLIP is trained with image-text pairs and lacks precise localization of objects while the image-level supervision has been used with heuristics that do not accurately specify local object regions. In this work, we propose to address this problem by performing object-centric alignment of …

Unpaired Style Transfer Conditional Generative Adversarial Network For Scanned Document Generation, David Jonathan Hawbaker

Dissertations and Theses

Neural networks are a powerful machine learning tool, especially when trained on a large dataset of relevant high-quality data. Generative adversarial networks, image super resolution and most other image manipulation neural networks require a dataset of images and matching target images for training. Collecting and compiling that data can be time consuming and expensive. This work explores an approach for building a dataset of paired document images with a matching scanned version of each document without physical printers or scanners. A dataset of these document image pairs could be used to train a generative adversarial network or image super resolution …

Fuzzy Reasoning Procedure For Ontologies Based On Rough Membership Approximation, Armand Florentin Donfack Kana, Babatunde Opeoluwa Akinkunmi

Future Computing and Informatics Journal

One of the major challenges in modeling a real-world domain is how to effectively represent uncertain and incomplete knowledge of that domain. Several techniques for representing uncertainty in ontologies have been proposed with some of the techniques lacking provision for vague inference. The classical tableaux-based algorithm does not provide the flexibility for reasoning over such vague ontologies. However, several extensions of the tableaux-based algorithm have been proposed to cope with fuzzy reasoning. Similarly, several alternative reasoning methods for incomplete, inconsistent, and uncertain ontologies have been proposed. One of the major limitations of most of those techniques is that they require …

Textual Emotion Detection Approaches: A Survey, Mahinda Mahmoud Samy Zidan, Ibrahim Elhenawy, Ahmed R. Abas, Mahmoud Othman

Future Computing and Informatics Journal

Over the past decades, social media attracted individuals to express their feelings on any topic or item, resulting in an incremental growth in the size of created data. These feelings and unstructured data paved the path for business organizations to gather information and build statistical analysis. Various machine learning and natural language processing-based approaches are used for sentiment and emotion analysis. Moreover, deep learning-based approaches recently gained popularity due to their remarkable performance in text analysis. This paper provides a comprehensive overview of the prominent machine learning models applied in emotion analysis. It explores various emotion analysis taxonomies, in addition …