Engineering Commons^™

Objective Estimation Of Tracheoesophageal Speech Quality, Yousef S Ettomi Ali

Electronic Thesis and Dissertation Repository

Speech quality estimation for pathological voices is becoming an increasingly important research topic. The assessment of the quality and the degree of severity of a disordered speech is important to the clinical treatment and rehabilitation of patients. In particular, patients who have undergone total laryngectomy (larynx removal) produce Tracheoesophageal (TE) speech. In this thesis, we study the problem of TE speech quality estimation using advanced signal processing approaches. Since it is not possible to have a reference (clean) signal corresponding to a given TE speech (disordered) signal, we investigate in particular the non-intrusive techniques (also called single-ended or blind approaches) …

Cluster-Based Chained Transfer Learning For Energy Forecasting With Big Data, Yifang Tian

Electronic Thesis and Dissertation Repository

Smart meter popularity has resulted in the ability to collect big energy data and has created opportunities for large-scale energy forecasting. Machine Learning (ML) techniques commonly used for forecasting, such as neural networks, involve computationally intensive training typically with data from a single building/group to predict future consumption for that same building/group. With hundreds of thousands of smart meters, it becomes impractical or even infeasible to individually train a model for each meter. Consequently, this paper proposes Cluster-Based Chained Transfer Learning (CBCTL), an approach for building neural network-based models for many meters by taking advantage of already trained models through …

Amodal Instance Segmentation And Multi-Object Tracking With Deep Pixel Embedding, Yanfeng Liu

Department of Electrical and Computer Engineering: Dissertations, Theses, and Student Research

This thesis extends upon the representational output of semantic instance segmentation by explicitly including both visible and occluded parts. A fully convolutional network is trained to produce consistent pixel-level embedding across two layers such that, when clustered, the results convey the full spatial extent and depth ordering of each instance. Results demonstrate that the network can accurately estimate complete masks in the presence of occlusion and outperform leading top-down bounding-box approaches.

The model is further extended to produce consistent pixel-level embeddings across two consecutive image frames from a video to simultaneously perform amodal instance segmentation and multi-object tracking. No post-processing …

Scatter Reduction By Exploiting Behaviour Of Convolutional Neural Networks In Frequency Domain, Carlos Ivan Jerez Gonzalez

Theses and Dissertations

In X-ray imaging, scattered radiation can produce a number of artifacts that greatly

undermine the image quality. There are hardware solutions, such as anti-scatter grids.

However, they are costly. A software-based solution is a better option because it is

cheaper and can achieve a higher scatter reduction. Most of the current software-based

approaches are model-based. The main issues with them are the lack of flexibility, expressivity, and the requirement of a model. In consideration of this, we decided to apply

Convolutional Neural Networks (CNNs), since they do not have any of the previously

mentioned issues.

In our approach we split …

Model Augmented Deep Neural Networks For Medical Image Reconstruction Problems, Hongquan Zuo

Theses and Dissertations

Solving an ill-posed inverse problem is difficult because it doesn't have a unique solution. In practice, for some important inverse problems, the conventional methods, e.g. ordinary least squares and iterative methods, cannot provide a good estimate. For example, for single image super-resolution and CT reconstruction, the results of these conventional methods cannot satisfy the requirements of these applications. While having more computational resources and high-quality data, researchers try to use machine-learning-based methods, especially deep learning to solve these ill-posed problems. In this dissertation, a model augmented recursive neural network is proposed as a general inverse problem method to solve these …

Machine Learning Methodology Review For Computational Electromagnetics, He Ming Yao, Lijun Jiang, Huan Huan Zhang, Wei E.I. Sha

Electrical and Computer Engineering Faculty Research & Creative Works

While machine learning is revolutionizing every corner of modern technologies, we have been attempting to explore whether machine learning methods could be used in computational electromagnetic (CEM). In this paper, five efforts in line with this direction are reviewed. They include forward methods such as the method of moments (MoM) solved by the artificial neural network training process, FDTD PML (perfectly matched layer) using the hyperbolic tangent basis function (HTBF), etc. There are also inverse problems that use the deep ConvNets for the effective source reconstruction and subwavelength imaging in the far-field. Benchmarks are provided to demonstrate the feasibility of …

Inverted Cone Convolutional Neural Network For Deboning Mris, Oliver John Palumbo

Theses and Dissertations

Data plenitude is the power but also the bottleneck for data-driven approaches, including neural networks. In particular, Convolutional Neural Networks (CNNs) require an abundant database of training images to achieve a desired high accuracy. Current techniques employed for boosting small datasets are data augmentation and synthetic data generation, which suffer from computational complexity and imprecision compared to original datasets. In this thesis, we intercalate prior knowledge based on the temporal relation between the images in the third dimension. Specifically, we compute the gradient of subsequent images in the dataset to remove extraneous information and highlight subtle variations between the images. …

Viewpoint Optimization For Autonomous Strawberry Harvesting With Deep Reinforcement Learning, Jonathon J. Sather

Master's Theses

Autonomous harvesting may provide a viable solution to mounting labor pressures in the United States' strawberry industry. However, due to bottlenecks in machine perception and economic viability, a profitable and commercially adopted strawberry harvesting system remains elusive. In this research, we explore the feasibility of using deep reinforcement learning to overcome these bottlenecks and develop a practical algorithm to address the sub-objective of viewpoint optimization, or the development of a control policy to direct a camera to favorable vantage points for autonomous harvesting. We evaluate the algorithm's performance in a custom, open-source simulated environment and observe affirmative results. Our trained …

High-Performance Learning Systems Using Low-Precision Nanoscale Devices, Nandakumar Sasidharan Rajalekshmi

Dissertations

Brain-inspired computation promises a paradigm shift in information processing, both in terms of its parallel processing architecture and the ability to learn to tackle problems deemed unsolvable by traditional algorithmic approaches. The computational capability of the human brain is believed to stem from an interconnected network of 100 billion compute nodes (neurons) that interact with each other through approximately 10¹⁵ adjustable memory junctions (synapses). The conductance of synapses is modifiable allowing the network to learn and perform various cognitive functions. Artificial neural networks inspired by this architecture have demonstrated even super-human performance in many complex tasks.

Computational systems based …

Machine Intelligence For Advanced Medical Data Analysis: Manifold Learning Approach, Fereshteh S Bashiri

Theses and Dissertations

In the current work, linear and non-linear manifold learning techniques, specifically Principle Component Analysis (PCA) and Laplacian Eigenmaps, are studied in detail. Their applications in medical image and shape analysis are investigated.

In the first contribution, a manifold learning-based multi-modal image registration technique is developed, which results in a unified intensity system through intensity transformation between the reference and sensed images. The transformation eliminates intensity variations in multi-modal medical scans and hence facilitates employing well-studied mono-modal registration techniques. The method can be used for registering multi-modal images with full and partial data.

Next, a manifold learning-based scale invariant global shape …

Receptive Fields Optimization In Deep Learning For Enhanced Interpretability, Diversity, And Resource Efficiency., Babajide Odunitan Ayinde

Electronic Theses and Dissertations

In both supervised and unsupervised learning settings, deep neural networks (DNNs) are known to perform hierarchical and discriminative representation of data. They are capable of automatically extracting excellent hierarchy of features from raw data without the need for manual feature engineering. Over the past few years, the general trend has been that DNNs have grown deeper and larger, amounting to huge number of final parameters and highly nonlinear cascade of features, thus improving the flexibility and accuracy of resulting models. In order to account for the scale, diversity and the difficulty of data DNNs learn from, the architectural complexity and …

Multi-Pig Part Detection And Association With A Fully-Convolutional Network, Eric T. Psota, Mateusz Mittek, Lance C. Pérez, Ty Schmidt, Benny Mote

Department of Electrical and Computer Engineering: Faculty Publications

Computer vision systems have the potential to provide automated, non-invasive monitoring of livestock animals, however, the lack of public datasets with well-defined targets and evaluation metrics presents a significant challenge for researchers. Consequently, existing solutions often focus on achieving task-specific objectives using relatively small, private datasets. This work introduces a new dataset and method for instance-level detection of multiple pigs in group-housed environments. The method uses a single fully-convolutional neural network to detect the location and orientation of each animal, where both body part locations and pairwise associations are represented in the image space. Accompanying this method is a new …

Deep Convolutional Neural Networks For Estimating Lens Distortion Parameters, Sebastian Lutz, Mark Davey, Aljosa Smolic

Session 2: Deep Learning for Computer Vision

In this paper we present a convolutional neural network (CNN) to predict multiple lens distortion parameters from a single input image. Unlike other methods, our network is suitable to create high resolution output as it directly estimates the parameters from the image which then can be used to rectify even very high resolution input images. As our method it is fully automatic, it is suitable for both casual creatives and professional artists. Our results show that our network accurately predicts the lens distortion parameters of high resolution images and corrects the distortions satisfactory.

Distribution Level Building Load Prediction Using Deep Learning, Abdulaziz S. Almalaq

Electronic Theses and Dissertations

Load prediction in distribution grids is an important means to improve energy supply scheduling, reduce the production cost, and support emission reduction. Determining accurate load predictions has become more crucial than ever as electrical load patterns are becoming increasingly complicated due to the versatility of the load profiles, the heterogeneity of individual load consumptions, and the variability of consumer-owned energy resources. However, despite the increase of smart grids technologies and energy conservation research, many challenges remain for accurate load prediction using existing methods. This dissertation investigates how to improve the accuracy of load predictions at the distribution level using artificial …

Lung Segmentation In Chest Radiographs Using Fully Convolutional Networks, Rahul Hooda, Ajay Mittal, Sanjeev Sofat

Turkish Journal of Electrical Engineering and Computer Sciences

Automated segmentation of medical images that aims at extracting anatomical boundaries is a fundamental step in any computer-aided diagnosis (CAD) system. Chest radiographic CAD systems, which are used to detect pulmonary diseases, first segment the lung field to precisely define the region-of-interest from which radiographic patterns are sought. In this paper, a deep learning-based method for segmenting lung fields from chest radiographs has been proposed. Several modifications in the fully convolutional network, which is used for segmenting natural images to date, have been attempted and evaluated to finally evolve a network fine-tuned for segmenting lung fields. The testing accuracy and …

Enhancing Face Pose Normalization With Deep Learning, Anil Çeli̇k, Nafi̇z Arica

Turkish Journal of Electrical Engineering and Computer Sciences

In this study, we propose a hybrid method for face pose normalization, which combines the 3-D model-based method with stacked denoising autoencoder (SDAE) deep network. Instead of applying a mirroring operation for the invisible face parts of the posed image, SDAE learns how to fill in those regions by a large set of training samples. In the performance evaluation, we compare the proposed method to four different pose normalization methods and investigate their effects on facial emotion recognition and verification problems in addition to visual quality tests. Methods evaluated in the experiments include 2-D alignment, 3-D model-based method, pure SDAE-based …

Novel Applications Of Machine Learning In Bioinformatics, Yi Zhang

Theses and Dissertations--Computer Science

Technological advances in next-generation sequencing and biomedical imaging have led to a rapid increase in biomedical data dimension and acquisition rate, which is challenging the conventional data analysis strategies. Modern machine learning techniques promise to leverage large data sets for finding hidden patterns within them, and for making accurate predictions. This dissertation aims to design novel machine learning-based models to transform biomedical big data into valuable biological insights. The research presented in this dissertation focuses on three bioinformatics domains: splice junction classification, gene regulatory network reconstruction, and lesion detection in mammograms.

A critical step in defining gene structures and mRNA …

Elimination Of Useless Images From Raw Camera-Trap Data, Ulaş Tekeli̇, Yalin Baştanlar

Turkish Journal of Electrical Engineering and Computer Sciences

Camera-traps are motion triggered cameras that are used to observe animals in nature. The number of images collected from camera-traps has increased significantly with the widening use of camera-traps thanks to advances in digital technology. A great workload is required for wild-life researchers to group and label these images. We propose a system to decrease the amount of time spent by the researchers by eliminating useless images from raw camera-trap data. These images are too bright, too dark, blurred, or they contain no animals. To eliminate bright, dark, and blurred images we employ techniques based on image histograms and fast …

Improving Word Embeddings Projection For Turkish Hypernym Extraction, Savaş Yildirim

Turkish Journal of Electrical Engineering and Computer Sciences

Corpus-driven approaches can automatically explore is-a relations between the word pairs from corpus. This problem is also called hypernym extraction. Formerly, lexico-syntactic patterns have been used to solve hypernym relations. The language-specific syntactic rules have been manually crafted to build the patterns. On the other hand, recent studies have applied distributional approaches to word semantics. They extracted the semantic relations relying on the idea that similar words share similar contexts. Former distributional approaches have applied one-hot bag-of-word (BOW) encoding. The dimensionality problem of BOW has been solved by various neural network approaches, which represent words in very short and dense …

Communications Using Deep Learning Techniques, Priti Gopal Pachpande

Legacy Theses & Dissertations (2009 - 2024)

Deep learning (DL) techniques have the potential of making communication systems

Multi-Column Neural Networks And Sparse Coding Novel Techniques In Machine Learning, Ammar O. Hoori

Theses and Dissertations

Accurate and fast machine learning (ML) algorithms are highly vital in artificial intelligence (AI) applications. In complex dataset problems, traditional ML methods such as radial basis function neural network (RBFN), sparse coding (SC) using dictionary learning, and particle swarm optimization (PSO) provide trivial results, large structure, slow training, and/or slow testing. This dissertation introduces four novel ML techniques: the multi-column RBFN network (MCRN), the projected dictionary learning algorithm (PDL) and the multi-column adaptive and non-adaptive particle swarm optimization techniques (MC-APSO and MC-PSO). These novel techniques provide efficient alternatives for traditional ML techniques. Compared to traditional ML techniques, the novel ML …