Computer Sciences Commons^™

Optimized Uncertainty Estimation For Vision Transformers: Enhancing Adversarial Robustness And Performance Using Selective Classification, Erik Pautsch, John Li, Silvio Rizzi, George K. Thiruvathukal, Maria Pantoja

Computer Science: Faculty Publications and Other Works

Deep Learning models often exhibit undue confidence when encountering out-of-distribution (OOD) inputs, misclassifying with high confidence. The ideal outcome, in these cases, would be an "I do not know" verdict. We enhance the trustworthiness of our models through selective classification, allowing the model to abstain from making predictions when facing uncertainty. Rather than a singular prediction, the model offers a prediction distribution, enabling users to gauge the model’s trustworthiness and determine the need for human intervention. We assess uncertainty in two baseline models: a Convolutional Neural Network (CNN) and a Vision Transformer (ViT). By leveraging these uncertainty values, we minimize …

Dtitd: An Intelligent Insider Threat Detection Framework Based On Digital Twin And Self-Attention Based Deep Learning Models, Zhi Qiang Wang, Abdulmotaleb El Saddik

Computer Vision Faculty Publications

Recent statistics and studies show that the loss generated by insider threats is much higher than that generated by external attacks. More and more organizations are investing in or purchasing insider threat detection systems to prevent insider risks. However, the accurate and timely detection of insider threats faces significant challenges. In this study, we proposed an intelligent insider threat detection framework based on Digital Twins and self-attentions based deep learning models. First, this paper introduces insider threats and the challenges in detecting them. Then this paper presents recent related works on solving insider threat detection problems and their limitations. Next, …

Deep Learning For Plant Bioinformatics: An Explainable Gradient-Based Approach For Disease Detection, Muhammad Shoaib, Babar Shah, Nasir Sayed, Farman Ali, Rafi Ullah, Irfan Hussain

All Works

Emerging in the realm of bioinformatics, plant bioinformatics integrates computational and statistical methods to study plant genomes, transcriptomes, and proteomes. With the introduction of high-throughput sequencing technologies and other omics data, the demand for automated methods to analyze and interpret these data has increased. We propose a novel explainable gradient-based approach EG-CNN model for both omics data and hyperspectral images to predict the type of attack on plants in this study. We gathered gene expression, metabolite, and hyperspectral image data from plants afflicted with four prevalent diseases: powdery mildew, rust, leaf spot, and blight. Our proposed EG-CNN model employs a …

Contemporary Art Authentication With Large-Scale Classification, Todd Dobbs, Abdullah-Al-Raihan Nayeem, Isaac Cho, Zbigniew Ras

Computer Science Faculty and Staff Publications

Art authentication is the process of identifying the artist who created a piece of artwork and is manifested through events of provenance, such as art gallery exhibitions and financial transactions. Art authentication has visual influence via the uniqueness of the artist’s style in contrast to the style of another artist. The significance of this contrast is proportional to the number of artists involved and the degree of uniqueness of an artist’s collection. This visual uniqueness of style can be captured in a mathematical model produced by a machine learning (ML) algorithm on painting images. Art authentication is not always possible …

Infrared Imaging Segmentation Employing An Explainable Deep Neural Network, Xinfei Liao, Dan Wang, Zairan Li, Nilanjan Dey, Rs Simon, Fuqian Shi

Turkish Journal of Electrical Engineering and Computer Sciences

Explainable AI (XAI) improved by a deep neural network (DNN) of a residual neural network (ResNet) and long short-term memory networks (LSTMs), termed XAIRL, is proposed for segmenting foot infrared imaging datasets. First, an infrared sensor imaging dataset is acquired by a foot infrared sensor imaging device and preprocessed. The infrared sensor image features are then defined and extracted with XAIRL being applied to segment the dataset. This paper compares and discusses our results with XAIRL. Evaluation indices are applied to perform various measurements for foot infrared image segmentation including accuracy, precision, recall, F1 score, intersection over union (IoU), Dice …

Cognitive Digital Modelling For Hyperspectral Image Classification Using Transfer Learning Model, Mohammad Shabaz, Mukesh Soni

Turkish Journal of Electrical Engineering and Computer Sciences

Deep convolutional neural networks can fully use the intrinsic relationship between features and improve the separability of hyperspectral images, which has received extensive in recent years. However, the need for a large number of labelled samples to train deep network models limits the application of such methods. The idea of transfer learning is introduced into remote sensing image classification to reduce the need for the number of labelled samples. In particular, the situation in which each class in the target picture only has one labelled sample is investigated. In the target domain, the number of training samples is enlarged by …

Affine Image Registration Of Arterial Spin Labeling Mri Using Deep Learning Networks, Zongpai Zhang, Huiyuan Yang, Yanchen Guo, Nicolas R. Bolo, Matcheri Keshavan, Eve Derosa, Adam K. Anderson, David C. Alsop, Lijun Yin, Weiying Dai

Computer Science Faculty Research & Creative Works

Convolutional neural networks (CNN) have demonstrated good accuracy and speed in spatially registering high signal-to-noise ratio (SNR) structural magnetic resonance imaging (sMRI) images. However, some functional magnetic resonance imaging (fMRI) images, e.g., those acquired from arterial spin labeling (ASL) perfusion fMRI, are of intrinsically low SNR and therefore the quality of registering ASL images using CNN is not clear. In this work, we aimed to explore the feasibility of a CNN-based affine registration network (ARN) for registration of low-SNR three-dimensional ASL perfusion image time series and compare its performance with that from the state-of-the-art statistical parametric mapping (SPM) algorithm. The …

An Ai-Based Framework For Translating American Sign Language To English And Vice Versa, Vijayendra D. Avina, Md Amiruzzaman, Stefanie Amiruzzaman, Linh B. Ngo, M. Ali Akber Dewan

Computer Science Faculty Publications

Abstract: In this paper, we propose a framework to convert American Sign Language (ASL) to English and English to ASL. Within this framework, we use a deep learning model along with the rolling average prediction that captures image frames from videos and classifies the signs from the image frames. The classified frames are then used to construct ASL words and sentences to support people with hearing impairments. We also use the same deep learning model to capture signs from the people with deaf symptoms and convert them into ASL words and English sentences. Based on this framework, we developed a …

Faster, Cheaper, And Better Cfd: A Case For Machine Learning To Augment Reynolds-Averaged Navier-Stokes, John Peter Romano Ii

Mechanical & Aerospace Engineering Theses & Dissertations

In recent years, the field of machine learning (ML) has made significant advances, particularly through applying deep learning (DL) algorithms and artificial intelligence (AI). The literature shows several ways that ML may enhance the power of computational fluid dynamics (CFD) to improve its solution accuracy, reduce the needed computational resources and reduce overall simulation cost. ML techniques have also expanded the understanding of underlying flow physics and improved data capture from experimental fluid dynamics.

This dissertation presents an in-depth literature review and discusses ways the field of fluid dynamics has leveraged ML modeling to date. The author selects and describes …

Object Recognition With Deep Neural Networks In Low-End Systems, Lillian Davis

Mahurin Honors College Capstone Experience/Thesis Projects

Object recognition is an important area in computer vision. Object recognition has been advanced significantly by deep learning that unifies feature extraction and classification. In general, deep neural networks, such as Convolution Neural Networks (CNNs), are trained in high-performance systems. Aiming to extend the reach of deep learning to personal computing, I propose a study of deep learning-based object recognition in low-end systems, such as laptops. This research includes how differing layer configurations and hyperparameter values used in CNNs can either create or resolve the issue of overfitting and affect final accuracy levels of object recognition systems. The main contribution …

Spoken Language Processing And Modeling For Aviation Communications, Aaron Van De Brook

Doctoral Dissertations and Master's Theses

With recent advances in machine learning and deep learning technologies and the creation of larger aviation-specific corpora, applying natural language processing technologies, especially those based on transformer neural networks, to aviation communications is becoming increasingly feasible. Previous work has focused on machine learning applications to natural language processing, such as N-grams and word lattices. This thesis experiments with a process for pretraining transformer-based language models on aviation English corpora and compare the effectiveness and performance of language models transfer learned from pretrained checkpoints and those trained from their base weight initializations (trained from scratch). The results suggest that transformer language …

Deep-Learning-Based Classification Of Digitally Modulated Signals, John A. Snoap

Electrical & Computer Engineering Theses & Dissertations

This dissertation presents several novel deep-learning (DL)-based approaches for classifying digitally modulated signals, one method of which involves the use of capsule networks (CAPs) together with cyclic cumulant (CC) features of the signals. These were blindly estimated using cyclostationary signal processing (CSP) and were then input into the CAP for training and classification. The classification performance and the generalization abilities of the proposed approach were tested using two distinct datasets that contained the same types of digitally modulated signals but had distinct generation parameters. The results showed that the classification of digitally modulated signals using CAPs and CCs proposed in …

Attention-Based Human Age Estimation From Face Images To Enhance Public Security, Md. Ashiqur Rahman, Shuhena S. Aonty, Kaushik Deb, Iqbal H. Sarker

Research outputs 2022 to 2026

Age estimation from facial images has gained significant attention due to its practical applications such as public security. However, one of the major challenges faced in this field is the limited availability of comprehensive training data. Moreover, due to the gradual nature of aging, similar-aged faces tend to share similarities despite their race, gender, or location. Recent studies on age estimation utilize convolutional neural networks (CNN), treating every facial region equally and disregarding potentially informative patches that contain age-specific details. Therefore, an attention module can be used to focus extra attention on important patches in the image. In this study, …

Perceptual Cue-Guided Adaptive Image Downscaling For Enhanced Semantic Segmentation On Large Document Images, Chulwoo Pack, Leen-Kiat Soh, Elizabeth Lorang

School of Computing: Faculty Publications

Image downscaling is an essential operation to reduce spatial complexity for various applications and is becoming increasingly important due to the growing number of solutions that rely on memory-intensive approaches, such as applying deep convolutional neural networks to semantic segmentation tasks on large images. Although conventional content-independent image downscaling can efficiently reduce complexity, it is vulnerable to losing perceptual details, which are important to preserve. Alternatively, existing content-aware downscaling severely distorts spatial structure and is not effectively applicable for segmentation tasks involving document images. In this paper, we propose a novel image downscaling approach that combines the strengths of both …

Style Transfer Network For Generating Opera Makeup Details, Fengquan Zhang, Duo Cao, Xiaohan Ma, Baijun Chen, Jiangxiao Zhang

Journal of System Simulation

Abstract: To address the problem of the loss of local style details in cross-domain image simulation, a ChinOperaGAN network framework suitable for opera makeup is designed from the perspective of protecting the excellent traditional culture. In order to solve the style translation of differences in two image domains, multiple overlapping local adversarial discriminators are proposed in the generative adversarial network. Since paired opera makeup data are difficult to obtain, a synthetic image is generated by combining the source image makeup mapping to effectively guide the transfer of local makeup details between images. In view of the characteristics of opera makeup …

A Deep Learning-Based Object Detection Framework For Automatic Asphalt Pavement Patch Detection Using Laser Profiling Images, Ibrahim Hassan Syed, Susan Mckeever Dr., Kieran Feighan, David Power, Dympna O'Sullivan

Conference papers

Road maintenance and the early detection of road defects rely on routine pavement inspections. While advanced 3D laser profiling systems have the capability to automatically identify certain types of distress such as cracks and ruts, more complex pavement damage, including patches, often require manual identification. To address this limitation, this study proposes an automated patch detection system that employs object detection techniques. The results demonstrate the ability of object detection models to accurately identify patches in laser profiling images, indicating that the proposed approach has the capability to significantly enhance automation in visual inspection processes. This has the potential for …

Towards Safe Automated Refactoring Of Imperative Deep Learning Programs To Graph Execution, Raffi Khatchadourian, Tatiana Castro Vélez, Mehdi Bagherzadeh, Nan Jia, Anita Raja

Publications and Research

Efficiency is essential to support responsiveness w.r.t. ever-growing datasets, especially for Deep Learning (DL) systems. DL frameworks have traditionally embraced deferred execution-style DL code—supporting symbolic, graph-based Deep Neural Network (DNN) computation. While scalable, such development is error-prone, non-intuitive, and difficult to debug. Consequently, more natural, imperative DL frameworks encouraging eager execution have emerged at the expense of run-time performance. Though hybrid approaches aim for the “best of both worlds,” using them effectively requires subtle considerations to make code amenable to safe, accurate, and efficient graph execution. We present our ongoing work on automated refactoring that assists developers in specifying whether …

Towards Safe Automated Refactoring Of Imperative Deep Learning Programs To Graph Execution, Raffi Khatchadourian, Tatiana Castro Vélez, Mehdi Bagherzadeh, Nan Jia, Anita Raja

Publications and Research

Efficiency is essential to support responsiveness w.r.t. ever-growing datasets, especially for Deep Learning (DL) systems. DL frameworks have traditionally embraced deferred execution-style DL code—supporting symbolic, graph-based Deep Neural Network (DNN) computation. While scalable, such development is error-prone, non-intuitive, and difficult to debug. Consequently, more natural, imperative DL frameworks encouraging eager execution have emerged at the expense of run-time performance. Though hybrid approaches aim for the "best of both worlds," using them effectively requires subtle considerations to make code amenable to safe, accurate, and efficient graph execution. We present our ongoing work on automated refactoring that assists developers in specifying whether …

Sentence Embedding Approach Using Lstm Auto-Encoder For Discussion Threads Summarization, Abdul Wali Khan, Feras Al-Obeidat, Afsheen Khalid, Adnan Amin, Fernando Moreira

All Works

Online discussion forums are repositories of valuable information where users interact and articulate their ideas and opinions, and share experiences about numerous topics. These online discussion forums are internet-based online communities where users can ask for help and find the solution to a problem. A new user of online discussion forums becomes exhausted from reading the significant number of irrelevant replies in a discussion. An automated discussion thread summarizing system (DTS) is necessary to create a candid view of the entire discussion of a query. Most of the previous approaches for automated DTS use the continuous bag of words (CBOW) …

Methods Of Evaluating Quantum Phase Estimation Circuit Output, Charles A. Woodrum

Theses and Dissertations

The quantum phase estimation (QPE) algorithm is one of the most important quantum computing algorithms that has been developed. The QPE algorithm estimates the phase or phases of the eigenvalue or eigenvalues of a unitary operator. It is a critical step for applications like Shor’s algorithm for factoring and the HHL algorithm for solving linear systems of equations, but it remains difficult to implement on current quantum computers due to small numbers of logical qubits and high error rates. This investigation derives a more accurate estimation of the phase of a unitary operator than would otherwise be attained with the …

Evaluation Of Novel Ai Architectures For Uncertainty Estimation, Erik Pautsch, John Li, Silvio Rizzi, George K. Thiruvathukal, Maria Pantoja

Computer Science: Faculty Publications and Other Works

Deep learning (DL) has become a cornerstone for advancements in computer vision, yielding models capable of remarkable performance on complex visual tasks. Despite these achievements, there remains a critical need for accurate uncertainty estimations, especially when models encounter out-of-distribution (OOD) inputs. Addressing this, our research focuses on the implementation and evaluation of uncertainty techniques in two prominent DL architectures: Convolutional Neural Networks (CNN) and Vision Transformers (ViT). These architectures were applied specifically to computer vision tasks, utilizing the MNIST and ImageNet-1K datasets for our evaluations.

High-Performance Computing (HPC) platforms, pivotal to this research, were employed to assess these techniques. The …

Arduinoprog: Towards Automating Arduino Programming, Imam Nur Bani Yusuf, Diyanah Binte Abdul Jamal, Lingxiao Jiang

Research Collection School Of Computing and Information Systems

Writing code for Arduino poses unique challenges. A developer 1) needs hardware-specific knowledge about the interface configuration between the Arduino controller and the I/Ohardware, 2) identifies a suitable driver library for the I/O hardware, and 3) follows certain usage patterns of the driver library in order to use them properly. In this work, based on a study of real-world user queries posted in the Arduino forum, we propose ArduinoProg to address such challenges. ArduinoProg consists of three components, i.e., Library Retriever, Configuration Classifier, and Pattern Generator. Given a query, Library Retriever retrieves library names relevant to the I/O hardware identified …

Experimental Comparison Of Features, Analyses, And Classifiers For Android Malware Detection, Lwin Khin Shar, Biniam Fisseha Demissie, Mariano Ceccato, Naing Tun Yan, David Lo, Lingxiao Jiang, Christoph Bienert

Research Collection School Of Computing and Information Systems

Android malware detection has been an active area of research. In the past decade, several machine learning-based approaches based on different types of features that may characterize Android malware behaviors have been proposed. The usually-analyzed features include API usages and sequences at various abstraction levels (e.g., class and package), extracted using static or dynamic analysis. Additionally, features that characterize permission uses, native API calls and reflection have also been analyzed. Initial works used conventional classifiers such as Random Forest to learn on those features. In recent years, deep learning-based classifiers such as Recurrent Neural Network have been explored. Considering various …

Rosas: Deep Semi-Supervised Anomaly Detection With Contamination-Resilient Continuous Supervision, Hongzuo Xu, Yijie Wang, Guansong Pang, Songlei Jian, Ning Liu, Yongjun Wang

Research Collection School Of Computing and Information Systems

Semi-supervised anomaly detection methods leverage a few anomaly examples to yield drastically improved performance compared to unsupervised models. However, they still suffer from two limitations: 1) unlabeled anomalies (i.e., anomaly contamination) may mislead the learning process when all the unlabeled data are employed as inliers for model training; 2) only discrete supervision information (such as binary or ordinal data labels) is exploited, which leads to suboptimal learning of anomaly scores that essentially take on a continuous distribution. Therefore, this paper proposes a novel semi-supervised anomaly detection method, which devises contamination-resilient continuous supervisory signals. Specifically, we propose a mass interpolation method …

On Explainability Of Neural Networks, Cem Benar

Dissertations

It is widely reported that deep neural networks outperform most competitors for a range of applications. The state-of-the-art neural networks have built-in inductive bias of architectural choices, regularizations, optimizer types, and initialization methods. Using inductive bias is intuitive to enhance the model approximation. Deep neural networks are mostly dense and heavily overparameterized. They tend to be biased towards low-rank solutions to reduce complexity and improve generalization performance, known as implicit regularization. The implicit regularization as observed in specific architectures and various real-world data sets suggests to overparameterize neural networks judiciously and learn compressed representations (lower rank approximation) with improved performance. …

A Proposed Artificial Intelligence Model For Android-Malware Detection, Fatma Taher, Omar Al Fandi, Mousa Al Kfairy, Hussam Al Hamadi, Saed Alrabaee

All Works

There are a variety of reasons why smartphones have grown so pervasive in our daily lives. While their benefits are undeniable, Android users must be vigilant against malicious apps. The goal of this study was to develop a broad framework for detecting Android malware using multiple deep learning classifiers; this framework was given the name DroidMDetection. To provide precise, dynamic, Android malware detection and clustering of different families of malware, the framework makes use of unique methodologies built based on deep learning and natural language processing (NLP) techniques. When compared to other similar works, DroidMDetection (1) uses API calls and …

Self-Supervised Pretraining And Transfer Learning On Fmri Data With Transformers, Sean Paulsen

Dartmouth College Ph.D Dissertations

Transfer learning is a machine learning technique founded on the idea that knowledge acquired by a model during “pretraining” on a source task can be transferred to the learning of a target task. Successful transfer learning can result in improved performance, faster convergence, and reduced demand for data. This technique is particularly desirable for the task of brain decoding in the domain of functional magnetic resonance imaging (fMRI), wherein even the most modern machine learning methods can struggle to decode labelled features of brain images. This challenge is due to the highly complex underlying signal, physical and neurological differences between …

Application Of Machine Learning Algorithms For Elucidation Of Biological Networks From Time Series Gene Expression Data, Krupa Nagori

Computational and Data Sciences (PhD) Dissertations

This dissertation provides a deep dive into understanding gene expression, interaction, regulation, and the intricate mechanisms behind heliotropism and phototropism. Additionally, the research accentuates the significance of machine learning techniques, specifically for gene regulatory networks (GRNs).

Chapter 1 offers an exhaustive benchmarking of GRN methodologies, furthering our comprehension of machine-learning models relevant to GRNs. The evaluation revealed that GRNTE, SWING, and BiXGBoost emerged as top-performing methods in GRN inference. The suitability of these models varies depending on specific research criteria such as computational needs, dataset dimensions, and performance metric emphasis. An innovation of this chapter was the introduction of Colab …

Seek And Classify: End-To-End Joint Multi-Signal Detection And Classification Using Deep Learning, Prashant Subedi

School of Computing: Dissertations, Theses, and Student Research

The rise in the use of wireless communication has led to the problem of spectrum scarcity in licensed bands. The popularity of Internet of Things (IoT) requires innovative solutions that maximize the use of available spectrum to support the increasing number of connected devices. This thesis tackles two significant problems in wireless communication: the need for efficient spectrum sensing techniques and the scarcity of large, diverse raw in-phase (I) and quadrature (Q) datasets.

The ability to detect and classify modulation of the signals efficiently can enable a cognitive radio to monitor the spectrum activity in real time and utilize unused …

Increasing The Efficiency And Accuracy Of Collective Intelligence Methods For Image Classification, Md Mahmudulla Hassan

Open Access Theses & Dissertations

Collective intelligence has emerged as a powerful methodology for annotating and classifying challenging data that pose difficulties for automated classifiers. It works by leveraging the concept of "wisdom of the crowds" which approximates a ground truth after aggregating experts' feedback and filtering out noise. However, challenges arise when certain applications, such as medical image classification, security threat detection, and financial fraud detection, demand accurate and reliable data annotation. The unreliability of experts due to inconsistent expertise and competencies, coupled with the associated cost and time-consuming judgment extraction, presents additional challenges.

Input aggregation is the process of consolidating and combining multiple …