Traditional Magnetic Resonance Imaging (MRI) reconstruction methods, which may be highly time-consuming and sensitive to noise, heavily depend on solving nonlinear optimization problems. By contrast, deep learning (DL)-based reconstruction methods do not need any explicit analytical data model and are robust to noise due to its large data-based training, which both make DL a versatile tool for fast and high-fidelity MR image reconstruction. While DL can be performed completely independently of traditional methods, it can, in fact, benefit from incorporating these established methods to achieve better results. To test this hypothesis, we proposed a hybrid DL-based MR image reconstruction method, which combines two state-of-the-art deep learning networks, U-Net and Generative Adversarial Network with Cycle loss (CycleGAN), with a traditional data reconstruction method: Projection Onto Convex Sets (POCS). Experiments were then performed to evaluate the method by comparing it to several existing state-of-the-art methods. Our results demonstrate that the proposed method outperformed the current state-of-the-art in terms of higher peak signal-to-noise ratio (PSNR) and higher Structural Similarity Index (SSIM). / Electrical and Computer Engineering

The abundance of next-generation sequencing (NGS) data has encouraged the adoption of machine learning methods to aid in the diagnosis and treatment of human disease. In particular, the last decade has shown the extensive use of predictive analytics in cancer research due to the prevalence of rich cellular descriptions of genetic and transcriptomic profiles of cancer cells. Despite the availability of wide-ranging forms of genomic data, few predictive models are designed to leverage multidimensional data sources. In this paper, we introduce a deep learning approach using neural network based information fusion to facilitate the integration of multi-platform genomic data, and the prediction of cancer cell sub-class. We propose the dGMU (deep gated multimodal unit), a series of multiplicative gates that can learn intermediate representations between multi-platform genomic data and improve cancer cell stratification. We also provide a framework for interpretable dimensionality reduction and assess several methods that visualize and explain the decisions of the underlying model. Experimental results on nine cancer types and four forms of NGS data (copy number variation, simple nucleotide variation, RNA expression, and miRNA expression) showed that the dGMU model improved the classification agreement of unimodal approaches and outperformed other fusion strategies in class accuracy. The results indicate that deep learning architectures based on multiplicative gates have the potential to expedite representation learning and knowledge integration in the study of cancer pathogenesis. / Thesis / Master of Science (MSc)

In this thesis we have present effective approaches for two classic Nature Language Processing tasks: Multi-label Text Classification(MLTC) and Sentiment Analysis(SA) based on two datasets. For MLTC, a robust deep learning approach based on convolution neural network(CNN) has been introduced. We have done this on almost one million records with a related label list consists of 20 labels. We have divided our data set into three parts, training set, validation set and test set. Our CNN based model achieved great result measured in F1 score. For SA, data set was more informative and well-structured compared with MLTC. A traditional word embedding method, Word2Vec was used for generating word vector of each text records. Following that, we employed several classic deep learning models such as Bi-LSTM, RCNN, Attention mechanism and CNN to extract sentiment features. In the next step, a classification frame was designed to graded. At last, the start-of-art language model, BERT which use transfer learning method was employed. In conclusion, we compared performance of RNN-based model, CNN-based model and pre-trained language model on classification task and discuss their applicability. / Thesis / Master of Science in Electrical and Computer Engineering (MSECE) / This theis purposed two deep learning solution to both multi-label classification problem and sentiment analysis problem.

El análisis de los movimientos oculares constituye un importante desafío dada la gran cantidad de información presente en los mismos. Estos movimientos proveen numerosas claves para estudiar diversos procesos cognitivos considerando, entre otros aspectos, el modo y el tiempo en que se codi fica la información y qué parte de los datos obtenidos se usan o se ignoran. Avanzar en el entendimiento de los procesos involucrados en tareas de alta carga cognitiva puede ayudar en la detección temprana de enfermedades neurodegenerativas tales como el mal de Alzheimer o el de Parkinson. A su vez, la comprensión de estos procesos puede ampliar el abordaje de una gran variedad de temas vinculados con el modelado y control del sistema oculomotor humano. Durante el desarrollo de esta Tesis Doctoral se llevaron a cabo tres experimentos que utilizan técnicas de deep-learning y modelos lineales de efecto mixto a n de identi car patrones de movimiento ocular a partir del estudio de situaciones controladas. La primera experiencia tiene como objetivo diferenciar adultos mayores sanos de adultos mayores con posible enfermedad de Alzheimer, utilizando deep-learning con denoise-sparse-autoencoders y un clasifi cador, a partir de información del movimiento ocular durante la lectura. Los resultados obtenidos, con un 89;8% de efectividad en la clasi ficación por oración y 100% por sujeto, son satisfactorios. Esto sugiere que el uso de esta técnica es una alternativa factible para esta tarea. La segunda experiencia tiene como objetivo demostrar la factibilidad de la utilización de la dilatación de la pupila como un marcador cognitivo, en este caso mediante modelos lineales de efecto mixto. Los resultados indican que la dilatación se ve influenciada por la carga cognitiva, la semántica y las características específi cas de la oración, por lo que representa una alternativa viable para el análisis cognitivo. El tercero y último experimento tiene como objetivo comprobar la efectividad de la utilización de redes neuronales recurrentes, con unidades LSTM, para lograr una clasifi cación efectiva en rangos etarios correspondientes a jóvenes sanos y adultos mayores sanos, a partir del análisis de la dinámica de la pupila. Los resultados obtenidos demuestran que la utilización de esta técnica tiene un alto potencial en este campo logrando clasifi car jóvenes vs. adultos mayores con una efectividad media por oración de 76;99% y una efectividad media por sujeto del 90;24 %, utilizando información del ojo derecho o información binocular. Los resultados de estos estudios permiten afi rmar que la utilización de técnicas de deep learning, que no han sido exploradas para resolver problemas como los planteados utilizando eye-tracking, constituyen un gran área de interés. / TEXTO PARCIAL en período de teletrabajo

Precision livestock farming technologies have been widely researched over the last decade. These technologies help in monitoring animal health and welfare parameters in a continuous, automated fashion. Under this umbrella of precision livestock farming, this study focuses on activity classification and body weight prediction in pigs. Activity monitoring is essential for understanding the health and growth of pigs. To automate this task effectively, we propose efficient and accurate sensor-based deep learning (DL) solutions. Among these, the 2D Residual Networks emerged as the best performing model, achieving an accuracy of 95.6%. This accuracy was 15.6% higher than that of other machine learning approaches. Additionally, accurate pig weight estimation is crucial for pork production, as it provides valuable insights into growth rates, disease prevalence, and overall health. Traditional manual methods of estimating pig weights are time-consuming and labor-intensive. To address this issue, we propose a novel approach that utilizes deep learning techniques on depth images for weight prediction. Through a custom image preprocessing pipeline, we train DL models to extract meaningful information from depth images for weight prediction. Our findings show that XceptionNet gives promising results, with a mean absolute error of 2.82 kg and a mean absolute percentage error of 7.42%. In comparison, the best performing statistical model, support vector machine, achieved a mean absolute error of 4.51 kg mean absolute percentage error of 15.56%. / Master of Science / With the increasing demand for food production in recent decades, the livestock farming industry faces significant pressure to modernize its methods. Traditional manual tasks such as activity monitoring and body weight measurement have been time-consuming and labor-intensive. Moreover, manual handling of animals can cause stress, negatively affecting their health. To address these challenges, this study proposes deep learning-based solutions for both activity classification and automated body weight prediction. For activity classification, our solution incorporates strategic data preprocessing techniques. Among various learning techniques, our deep learning model, the 2D Residual Networks, achieved an accuracy of 95.6%, surpassing other approaches by 15.6%. Furthermore, this study also compares statistical models with deep learning models for the body weight prediction task. Our analysis demonstrates that deep learning models outperform statistical models in terms of accuracy and inference time. Specifically, XceptionNet yielded promising results, with a mean absolute error of 2.82 kg and a mean absolute percentage error of 7.42%, outperforming the best statistical model by nearly 8%.

Modern deep neural networks do not produce well calibrated estimates of their own uncertainty, unless specific uncertainty estimation techniques are applied. Common uncertainty estimation techniques such as Deep Ensembles and Monte Carlo Dropout necessitate multiple forward pass evaluations for each input sample, making them too slow for real-time use. For real-time use, techniques which require only a single-forward pass are desired. Evidential Deep Learning (EDL), and Multiple-Input Multiple-Output (MIMO) networks are prior art in the space of real-time uncertainty estimation. This work introduces EDL-MIMO, a novel real-time uncertainty estimation method which combines the two. The core of this thesis is dedicated to comparing the quality of this new method to the pre-existing baselines of EDL and MIMO alone. / De neurala nätverk vi har idag har svårigheter med att bedöma sin egen osäkerhet utan särskilda metoder. Metoder som Deep Ensembles och Monte Carlo Dropout kräver flera beräkningar för varje indata, vilket gör dem för långsamma i realtid. För realtidstillämpning behövs metoder som endast kräver en beräkning. Det finns redan vetenskapliga artiklar om osäkerhetsmetoder som Evidential Deep Learning (EDL), och Multiple-Input Multiple-Output (MIMO) networks. Denna uppsats introducerar en ny metod som kombinerar båda. Fokus ligger på att jämföra kvaliteten på denna nya metod med EDL och MIMO när de används ensamma / Djúptauganet nútímans eiga erfitt með að meta sína eigin óvissu, án þess að sérstakar óvissumatsaðferðir séu notaðar. Algengar óvissumatsaðferðir líkt og Deep Ensembles, og Monte Carlo Dropout, krefjast þess að djúptauganetið sé reiknað oftar en einu sinni fyrir hvert inntak, sem gerir þessar aðferðir of hægar fyrir rauntímanotkun. Fyrir rauntímanotkun er leitast eftir aðferðum sem krefjast bara einn reikning. Evidential Deep Learning (EDL), og Multiple-Input Multiple-Output (MIMO) networks eru óvissumatsaðferðir sem hafa verið birtar í fyrri greinum. Þessi ritgerð kynnir í fyrsta sinn EDL-MIMO, nýja óvissumatsaðferð sem blandar þeim báðum saman. Kjarni þessarar ritgerðar snýst um að bera saman gæði þessarar nýju aðferðar í samanburð við að nota EDL eða MIMO einar og sér.

Yes / Today, computer based face recognition is a mature and reliable mechanism which is being practically utilised for many access control scenarios. As such, face recognition or authentication is predominantly performed using ‘perfect’ data of full frontal facial images. Though that may be the case, in reality, there are numerous situations where full frontal faces may not be available — the imperfect face images that often come from CCTV cameras do demonstrate the case in point. Hence, the problem of computer based face recognition using partial facial data as probes is still largely an unexplored area of research. Given that humans and computers perform face recognition and authentication inherently differently, it must be interesting as well as intriguing to understand how a computer favours various parts of the face when presented to the challenges of face recognition. In this work, we explore the question that surrounds the idea of face recognition using partial facial data. We explore it by applying novel experiments to test the performance of machine learning using partial faces and other manipulations on face images such as rotation and zooming, which we use as training and recognition cues. In particular, we study the rate of recognition subject to the various parts of the face such as the eyes, mouth, nose and the cheek. We also study the effect of face recognition subject to facial rotation as well as the effect of recognition subject to zooming out of the facial images. Our experiments are based on using the state of the art convolutional neural network based architecture along with the pre-trained VGG-Face model through which we extract features for machine learning. We then use two classifiers namely the cosine similarity and the linear support vector machines to test the recognition rates. We ran our experiments on two publicly available datasets namely, the controlled Brazilian FEI and the uncontrolled LFW dataset. Our results show that individual parts of the face such as the eyes, nose and the cheeks have low recognition rates though the rate of recognition quickly goes up when individual parts of the face in combined form are presented as probes.

Yes / Event monitoring is an essential application of Smart City platforms. Real-time monitoring of gully and drainage blockage is an important part of flood monitoring applications. Building viable IoT sensors for detecting blockage is a complex task due to the limitations of deploying such sensors in situ. Image classification with deep learning is a potential alternative solution. However, there are no image datasets of gullies and drainages. We were faced with such challenges as part of developing a flood monitoring application in a European Union-funded project. To address these issues, we propose a novel image classification approach based on deep learning with an IoT-enabled camera to monitor gullies and drainages. This approach utilises deep learning to develop an effective image classification model to classify blockage images into different class labels based on the severity. In order to handle the complexity of video-based images, and subsequent poor classification accuracy of the model, we have carried out experiments with the removal of image edges by applying image cropping. The process of cropping in our proposed experimentation is aimed to concentrate only on the regions of interest within images, hence leaving out some proportion of image edges. An image dataset from crowd-sourced publicly accessible images has been curated to train and test the proposed model. For validation, model accuracies were compared considering model with and without image cropping. The cropping-based image classification showed improvement in the classification accuracy. This paper outlines the lessons from our experimentation that have a wider impact on many similar use cases involving IoT-based cameras as part of smart city event monitoring platforms. / European Regional Development Fund Interreg project Smart Cities and Open Data REuse (SCORE).

Yes / Assuring safety and thereby certifying is a key challenge of many kinds of Machine Learning (ML) Models. ML is one of the most widely used technological solutions to automate complex tasks such as autonomous driving, traffic sign recognition, lane keep assist etc. The application of ML is making a significant contributions in the automotive industry, it introduces concerns related to the safety and security of these systems. ML models should be robust and reliable throughout and prove their trustworthiness in all use cases associated with vehicle operation. Proving confidence in the safety and security of ML-based systems and there by giving assurance to regulators, the certification authorities, and other stakeholders is an important task. This paper proposes a framework to handle uncertainties of ML model to improve the safety level and thereby certify the ML Models in the automotive industry.

<p dir="ltr">Cerebrospinal fluid (CSF) plays a diverse role within the skull including cushioning the brain, regulating intracranial pressure, and clearing metabolic waste via the glymphatic system. Disruptions in CSF flow have long been investigated for hydrocephalus-related diseases such as idiopathic normal pressure hydrocephalus (iNPH). Recently, changes in CSF flow have been implicated in neurodegenerative disorders such as Alzheimer’s disease (AD) and Parkinson’s disease. It remains difficult to obtain <i>in vivo </i>measurements of CSF flow which contribute to disease initiation, progression, and treatment. Three-directional phase-contrast MR imaging (4D flow MRI) has been used to measure CSF velocities within the cerebral ventricles. However, there remain challenges in balancing acquisition time, spatiotemporal resolution, and velocity-to-noise ratio. This is complicated by the low velocities and long relaxation times associated with CSF flow. Additionally, flow-derived metrics associated with cellular adaptations and transport rely on near-wall velocities which are poorly resolved and noisy. To address these challenges, we have applied physics-guided neural networks (PGNN) to super-resolve and denoise synthetic 4D flow MRI of CSF flow within the 3rd and 4th ventricles using novel physics-based loss functions. These loss functions are specifically designed to ensure that high-resolution estimations of flow fields are physically consistent and temporarily coherent. We apply these PGNN to various test cases including synthetically generated 4D flow MRI in the cerebral ventricles and vasculature, <i>in vitro</i> 4D flow MRI acquired at two resolutions in 3D printed phantoms of the 3rd and 4th ventricles, and in vivo 4D flow MRI in a healthy subject. Lastly, we apply these physics-guided networks to investigate blood flow through cerebral aneurysms. These techniques can empower larger studies investigating the coupling between arterial blood flow and CSF flow in conditions such as iNPH and AD.</p>