Filetype
Epilepsy is a neurological disorder, affecting approximately 1% of the population. The high rate of drug-resistance in epilepsy in general (30%), and temporal lobe epilepsy (TLE) in particular, pose a serious clinical problem. Thus, novel treatments for drug-resistant epilepsy is highly warranted. Gene therapy has been suggested to be a promising approach to target drug-resistant focal epilepsy si
The problem of monotone smoothing splines with bounds is formulated as a constrained minimization problem of the calculus of variations. Existence and uniqueness of solutions of this problem is proved, as well as the equivalence of it to a finite dimensional but nonlinear optimization problem. A new algorithm for computing the solution which is a spline curve, using a branch and bound technique, i
Background: Demography, politics, economy, and governance appear to be the major structural factors for health and well-being. These factors have a significant role to play in achieving universal health coverage (UHC). The majority of previous studies did not highlight those factors. The aim of this study is to explore the basic structural factors (political stability, demography, gross national i
We compare classic scalar temporal difference learning with three new distributional algorithms for playing the game of 5-in-a-row using deep neural networks: distributional temporal difference learning with constant learning rate, and two distributional temporal difference algorithms with adaptive learning rate. All these algorithms are applicable to any two-player deterministic zero sum game and
Pharmaceuticals that are not completely metabolized in the human body are excreted in wastewater and may find their way into surface waters, where they can have harmful effects on aquatic fauna and flora. The implementation of ozonation at wastewater treatment plants (WWTPs) can reduce the amount of pharmaceuticals discharged to the environment. Despite the technology readiness level of ozone trea
We present a model for generating 3d articulated pedestrian locomotion in urban scenarios, with synthesis capabilities informed by the 3d scene semantics and geometry. We reformulate pedestrian trajectory forecasting as a structured reinforcement learning (RL) problem. This allows us to naturally combine prior knowledge on collision avoidance, 3d human motion capture and the motion of pedestrians
3D human pose and shape estimation plays a vital role in many computer vision applications. There are many deep learning based methods attempting to solve the problem only relying on single-view RGB images for training the network. However, since some public datasets are captured from multi-view cameras system, we propose a novel method to tackle the problem by putting optimization-based multi-vie
3D human body reconstruction from monocular images has wide applications in our life, such as movie, animation, Virtual/Augmented Reality, medical research and so on. Due to the high freedom of human body in real scene and the ambiguity of inferring 3D objects from 2D images, it is a challenging task to accurately recover 3D human body models from images. In this thesis, we explore the methods for
A wide range of medical examinations are using analysis of images from different types of equipment. Using artificial intelligence, the assessments could be done automatically. This can have multiple benefits for the healthcare; reduce workload for medical doctors, decrease variations in diagnoses and cut waiting times for the patient as well as improve the performance. The aim of this thesis has
Nondestructive imaging with both a large field of view and a high spatial resolution is crucial to understand complex materials and processes in science and technology. X-ray ptychography can provide highest spatial resolution but is limited in the field of view by the acquisition time and coherent flux at modern x-ray sources. By multi-beam ptychography, the sample can be imaged in parallel by se
Regression via classification (RvC) is a common method used for regression problems in deep learning, where the target variable belongs to a set of continuous values. By discretizing the target into a set of non-overlapping classes, it has been shown that training a classifier can improve neural network accuracy compared to using a standard regression approach. However, it is not clear how the set
Orientation towards application, transdisciplinarity and social distribution are prominent ideas when policy prescribes how to conduct research on complex, real-world issues. Issues dealt with in ageing research are both epistemologically complex and politically contentious. By using the metaphor of ‘travelling ideas,' we have followed these ideas on their journey into 17 approved grant proposals
Modelling capabilities have drastically improved in the last decade. However, in most of the cases the fire response of building elements is predicted by fitting input material properties to the models in order to match test data. This paper presents models developed to predict the unexposed side temperature of stone wool layered composites with stainless steel or gypsum claddings exposed to sever
The scan-on-receive (SCORE) is a key digital beamforming (DBF) technique for future high-resolution wide-swath synthetic aperture radar (SAR) systems. A sharp and high gain receive beam, steered in real time towards the expected direction of arrival (DoA) of the backscattering signal, allows for improved SAR imaging performance with respect to a conventional approach. Nevertheless, it also exposes
