The photocathode is one of the key components of particle accelerator facilities that provides electrons for experiments in many disciplines such as biomedicine, security imaging, and condensed matter physics. The requirements for the electron emitting material, the so-called photocathode, are rather high because these materials should provide a high quantum efficiency, a low thermal emittance, a fast response, and a long operational lifetime. At present, none of the state-of-the-art photocathodes can fully meet all the desired requirements. Therefore, new materials that can be used as potential photocathodes are urgently needed for future developments in accelerator research. Semiconductor photocathodes such as cesium telluride are the preferred materials in particle accelerators. These photocathodes provide high quantum efficiencies of above 10 %, making them highly attractive. The crystal growth of cesium telluride, as a compound semiconductor photocathode, requires the deposition of cesium and tellurium on a suitable substrate with an ideal chemical ratio, which seems elaborate and difficult to handle. In contrast, III-V semiconductors, such as gallium arsenide and gallium nitride (GaN), represent another type of semiconductor photocathode. These commercially available semiconductors are already grown on a substrate and only require a thin film of cesium and optional oxygen to obtain a photocathode. An atomically clean surface is necessary to achieve a negative electron affinity surface, which is the main prerequisite for high quantum efficiency. In this work, p-GaN grown on sapphire by metal-organic chemical vapor deposition, was wet chemically cleaned, and transferred into an ultra-high vacuum chamber, where it underwent a subsequent thermal cleaning. The cleaned p-GaN samples were activated with Cs to obtain p-GaN:Cs photocathodes and their performance was monitored with respect to their quality, especially concerning their quantum efficiency and storage lifetime. The surface topography and morphology were examined ex-situ by atomic force microscopy and scanning electron microscopy in combination with energy dispersive X-ray spectroscopy. Treatments at different temperatures resulted in various quantum efficiency values and storage lifetimes. Moderate temperatures of 400–500 °C were found to be more beneficial for the p-GaN surface quality, which was reflected by achieving higher quantum efficiency values. After the thermal cleaning, the samples were activated with a thin layer of cesium at an average pressure of 1 x E-9 mbar. The surface morphology was studied with scanning electron microscopy and energy dispersive X-ray spectroscopy after the samples were thermally cleaned and activated with cesium. The results showed that the surface appeared inhomogeneous when the samples were cleaned at a high temperature above 600 °C. A thermal cleaning from the back side through the substrate represented another possibility but did not yield higher quantum efficiency values. An in-situ analysis method facilitates following and understanding the changes in the surface electronic states before, during, and after any treatment of p-GaN:Cs photocathodes. For this purpose, an X-ray photoelectron spectrometer was applied that was built into an ultra-high vacuum system to prepare and characterize photocathodes. It allowed the in-situ monitoring of the photocathode surfaces beginning immediately after their cleaning and throughout the activation and degradation processes. The realization of the adaption of an X-ray photoelectron spectroscopy chamber to the preparation chamber presented a significant constructional challenge. Thus, this work paid special attention to the technical aspects of in-situ sample transportation between these chambers without leaving the ultra-high vacuum environment. The p-GaN surface was cleaned with different solutions and studied by X-ray photoelectron spectroscopy and atomic force microscopy, revealing that cleaning with a so-called 'piranha' solution in combination with rinsing in ethanol works best for the p-GaN surface. A cleaning step that solely uses ethanol is also possible and represents a simple cleaning procedure that is manageable in all laboratories. Afterward, the cleaned p-GaN samples underwent a subsequential thermal vacuum cleaning at various temperatures to achieve an atomically clean surface. Each treatment step was followed by X-ray photoelectron spectroscopy analysis without leaving the ultra-high vacuum environment, revealing residual oxygen and carbon on the p- GaN surface. A thermal treatment under vacuum did not entirely remove these organic contaminations, although the thermal cleaning reduced their peak intensities. The remaining oxygen and carbon contaminants were assumed to be residuals derived from the metal-organic chemical vapor deposition process. After the cesium activation, a shift toward a higher binding energy was observed in the X-ray photoelectron spectroscopy spectra of the related photoemission peaks. This shift indicated that the cesium was successfully adsorbed to the p-GaN surface. Before the cesium activation, adventitious carbon at a binding energy of approximately 284 eV was found, which was also present after the cesium activation but did not shift in its binding energy. It was also shown that the presence of remaining carbon significantly influenced the photocathode’s quality. After the cesium deposition, a new carbon species at a higher binding energy (approximately 286 eV) appeared in the carbon 1s spectrum. This new species showed a higher binding energy than adventitious carbon and was identified as a cesium carbide species. This cesium carbide species grew over time, resulting in islands on the surface. The X-ray photoelectron spectroscopy data facilitated the elucidation of the critical role of thiscesium carbide species in photocathode degradation. Typically, the quantum efficiency of photocathodes decays exponentially. Conversely, an immense quantum efficiency loss was observed after the p-GaN:Cs photocathodes were studied by X-ray photoelectron spectroscopy. The origin of the quantum efficiency loss derived from X-rays as an external influence and was not caused by the sample’s transportation. Therefore, potential X-ray damages to the p-GaN:Cs photocathodes were investigated. These experiments showed that the adsorbed cesium and its adhesion to the p-GaN surface were strongly influenced by X-ray irradiation. The cesium photoemission peaks shifted toward a lower binding energy, while the relative cesium concentration did not. This shift indicated that X-ray irradiation accelerated the external aging of the p-GaN photocathodes and thus it was proposed to use lower X-ray beam power or cool the samples to prevent X-ray damage to cesiated photocathodes. This work shows that an exclusive activation with cesium is feasible and that a re-activation of the same sample is possible. Quantum efficiency values of 1–12% were achieved when the p-GaN, grown on sapphire, was activated. The capability of an X-ray photoelectron spectroscopy analysis allowed the in-situ monitoring of the photocathode surface and shed light on the surface compositions that changed during the photocathodes’ degradation process.

The power of generative deep learning has increased very quickly in the past ten years and modern models are now able to generate human faces that are indistinguishable from real ones. This thesis project will investigate the uses and limitations of this technology by attempting to generate very specific data, images of golf holes. Generative adverserial networks, GANs, were used to solve this problem. Two different GAN models were chosen as candidates and these were trained on some different datasets that were extracted from the project provider Topgolf Sweden AB’s virtual golf game. This golf game contained data of many different types of golf holes from all over the world. The best performing model was Progressive Growing GAN, ProGAN, which works by iteratively increasing the size of the images until the desired size is reached. This model was able to produce results of very high quality and with large variety. To further investigate the quality of the results a survey was sent out to the employees of Topgolf Sweden AB. A survey that showed that it was difficult for the participants to correctly determine if a given image was real or had been generated by the model. These results further showed that the generated samples had a high quality. This thesis project also investigated how height data could be incorporated in the process. The results showed that the ProGAN model was able to generate height maps that capture the most important aspects of a golf hole. Furthermore, the overall results showed that the generative model had learned a good representation of the data’s underlying probability distribution. More work needs to be done before a model like the one presented here can be used to generate complete golf holes that can be used in a virtual golf game, but this project clearly shows that GANs are a worthwhile investment for this purpose. / Kraften i generativ djupinlärning har ökat snabbt under de senaste tio åren och moderna modeller kan generera bilder på människoansikten som är omöjliga att urskilja från riktiga ansikten. Detta examensarbete undersöker hur denna teknologi kan användas och vad det finns för begränsningar genom att försöka generera väldigt specifik data, bilder på golfhål. Generativa adversiella nätverk, GANs, användas för att lösa detta problem. Två modeller valdes som kandidater och dessa tränades på olika datasets som hade extraherats från projektleverantören Topgolf Sweden ABs virtuella golfspel. Detta golfspel innehöll data från en mängd olika typer av golfhål från hela världen. Modellen som presterade bäst var Progressive Growing GAN, ProGAN, som iterativt ökar storleken på bilderna tills den önskade storleken har nåtts. Denna modell lyckades skapa bilder av väldigt hög kvalitet och med stor variation. För att ytterligare undersöka kvaliten på resultaten så genomfördes en enkät. Enkäten skickades till anställda hos Topgolf Sweden AB. Svaren visade att det var svårt för deltagarna att urskilja äkta bilder från genererade bilder vilket ytterligare visade att de genererade bilderna hade hög kvalitet. Detta examensarbete undersökte också hur höjddata kunde integreras i processen. Resultaten av detta visade att ProGAN modellen kunde generera höjddata som innehöll de viktigaste delarna av ett golfhål. Dessutom så visade resultaten i helhet att den generativa modellen hade lärt sig en bra representation av träningsdatans underliggande sannolikhetsfördelning. Mer arbete krävs för att en liknande modell ska kunna generera kompletta golfhål som kan användas i ett virtuellt golfspel, men projektet visar att GANs är ett väldigt bra alternativ för att lyckas med det.

The TRACAB Image Tracking System is used by ChyronHego for the tracking of ball and players on football fields. It requires the calibration of the cameras around the arena which is disrupted by fences and other mesh structures that are positioned between the camera and the field as a safety measure for the public. The purpose of this work was the implementation of a cycle consistent Generative Adversarial Network (cycle-GAN) for removing the fence from the image using unpaired data. Cycle-GANs are part of the state-of-the-art of image-to-image translation and can solve this kind of problem without the need of paired images. This makes it an exciting and powerful method and, according to the latest investigations in the current work, it has never been used for this kind of application before. The model was able to strongly attenuate, and in some cases completely remove, the net structure from images. To quantify the impact of the net removal a homography matching was performed. Then, it was compared with the homography associated to the baseline of blurring the image with a gaussian filter and the original image without the use of any filter. The results showed that the identification of key-points was harder on synthetic images than on the original image with or without small Gaussian filters, but it showed a better performance against images blurred with filters with a standard deviation of 3 pixels or more. Despite the performance not being better than the baseline in all the cases it always added new key-points, and sometimes, it was able to find correct homographies where the baseline could not. Therefore, the cycle-GAN model proved to complement the baseline. / TRACAB Image Tracking System används av ChyronHego för spårning av bollen och spelaren påfotbollsplaner. Detta kräver kalibrering av kamerorna runt arenan som störs av staket och andra nätstrukturer som är placerade mellan kameran och fältet som en säkerhetsåtgärd för publiken. Detta examensabrete fokuserar påimplementeringen av en cycle-GAN för borttagning av nätet från bilden med hjälp av oparade data. Cycle-GAN är en bild-till-bild-översättning state-of-the-art teknik och det kan lösa denna typ av problem utan parade bilder. Detta gör det till en spännande och kraftfull metod och enligt den senaste forskningen har det aldrig använts för denna typ av tillämpning förut. Modellen kunde kraftigt dämpa och i vissa fall helt ta bort nätstrukturen från bilder. För att kvantifiera effekterna av avlägsnandet av nätet utfördes en homografimatchning. Därefter jämfördes det med homografin associerad med baslinjen där bilden görs suddig med ett gaussiskt filter och originalbilden utan användning av något filter. Resultaten visade att identifieringen av nyckelpunkter var svårare påsyntetiska bilder än påoriginalbilder med eller utan småGauss-filter, men det visade bättre prestanda än bilder som var suddigt med filter med en standardavvikelse på 3 pixlar eller mer. Trots att prestandan inte var bättre än baslinjen i alla fall lade versionen utan nätet alltid till nya nyckelpunkter, och ibland kunde den hitta korrekta homografier där baslinjen misslyckades. Därför, cycle-GAN-modellen kompletterar baslinjen.

Anomaly detection is a field of study that is closely associated with machine learning and it is the process of finding irregularities in datasets. Developing and maintaining multiple machine learning models for anomaly detection takes time and can be an expensive task. One proposed solution is to combine all datasets and create a single model. This creates a heterogeneous dataset with a wide variation in its distribution, making it difficult to find anomalies in the dataset. The objective of this thesis is then to identify a framework that is suitable for anomaly detection in heterogeneous datasets. A selection of five methods were implemented in this project - 2 supervised learning approaches and 3 unsupervised learning approaches. These models are trained on 3 synthetic datasets that have been designed to be heterogeneous with an imbalance between the classes as anomalies are rare events. The performance of the models are evaluated with the AUC and the F1-score, aswell as observing the Precision-Recall Curve. The results makes it evident that anomaly detection in heterogeneous datasets is a challenging task. The best performing approach was with a random forest model where the class imbalance problem had been solved by generating synthetic samples of the anomaly class by implementing a generative adversarial network. / Anomalidetektering är ett studieområde som är starkt förknippat med maskininlärning och det kan beskrivas som processen att hitta avvikelser i datamängder. Att utveckla och underhålla flera maskininlärningsmodeller tar tid och kan vara kostsamt. Ett förslag för att lösa dessa problem är att kombinera alla dataset och skapa endast en modell. Detta leder till att datamängden blir heterogen i dess fördelning och gör det mer utmanande att skapa en modell som kan detektera anomalier. Syftet i denna tes är att identifiera ett ramverk som är lämpligt för anomalidetektering i heterogena datamängder. Ett urval av fem metoder tillämpades i detta projekt - 2 metoder inom övervakad inlärning och 3 metoder inom oövervakad inlärning. Dessa modeller är tränade på syntetiska datamängder som är framtagna så att de är heterogena i dess fördelning och har en urbalans mellan klasserna då anomalier är sällsynta händelser. Modellernas prestanda evalueras genom att beräkna dess AUC och F1-värde, samt observera Precision-Recall kurvan. Resultaten gör det tydligt att anomalidetektering i heterogena datamängder är ett utmanande uppdrag. Den model som presterade bäst var en random forest model där urbalansen mellan klasserna var omhändertagen genom att generera syntetiska observation av anomaliklassen med hjälp av en generativ advarserial network.

GaN-HEMTs (Gallium Nitride-based High Electron Mobility Transistors) have, thanks to the large band gap of GaN, electrical properties that are suitable for applications of high electrical voltages, high currents, and fast switching. The large band gap also gives GaN-HEMTs a high resistance to radiation. In this degree project, the effects of 2 MeV proton irradiation of GaN-HEMTs constructed on both silicon carbide and silicon substrates are investigated. 20 transistors per substrate were irradiated in the particle accelerator 5 MV NEC Pelletron in the Ångström laboratory at Uppsala University. These transistors were exposed to radiation doses in the range of 10^11 to 10^15 protons/cm^2. The analysis shows that both transistors on silicon, as well as silicon carbide, are unaffected by proton irradiation up to a dose of 10^14 protons/cm^2. GaN-on-Si transistors show less influence of radiation than GaN-on-SiC transistors. The capacitances between gate and drain as well as drain and source for both GaN-on-SiC and GaN-on-Si HEMTs show hysteresis as a function of forward and backward gate voltage sweeps for the radiation dose of 10^15 protons/cm^2. / GaN-HEMTs (Galliumnitridbaserade High Electron Mobility Transistors) har tack vare det stora bandgapet i GaN goda elektriska egenskaper som lämpar sig för höga elektriska spänningar, höga strömmar och snabb växling mellan av- och på-tillstånd. Det stora bandgapet ger även GaN-HEMTs ett stort motstånd mot strålning.I detta examensarbete undersöks effekterna av 2 MeV protonbestrålning av GaN-HEMTs. Dessa HEMTs är konstruerade på både kiselkarbid- och kiselsubstrat.20 transistorer per transistorsubstrat bestrålades i partikelacceleratorn 5 MV NEC Pelletron i Ångströmslaboratoriet vid Uppsala Universitet. Dessa transistorer utsattes för strålningsdoser inom intervallet 10^11 till 10^15 protoner/cm^2. Resultaten visar att både tranisistorer på kisel såsom kiselkarbid är opåverkade av strålning upp till en dos av 10^14 protoner/cm^2. GaN-på-Si-transistorer visar en mindre påverkan av protonstrålning än GaN-på-SiC-transistorer. Ytterligare uppstod hysteresis för kapacitanser mellan gate och drain och mellan gate och source som en funktion av fram- och bakriktad gate-spänning efter en strålningsdos av 10^15 protoner/cm^2.

Following the introduction of new laws and regulations to ensure data protection in GDPR and PIPEDA, interests in technologies to protect data privacy have increased. A promising research trajectory in this area is found in Generative Adversarial Networks (GAN), an architecture trained to produce data that reflects the statistical properties of its underlying dataset without compromising the integrity of the data subjects. Despite the technology’s young age, prior research has made significant progress in the generation process of so-called synthetic data, and the current models can generate images with high-quality. Due to the architecture’s success with images, it has been adapted to new domains, and this study examines its potential to synthesize financial tabular data. The study investigates a state-of-the-art model within tabular GANs, called CTGAN, together with two proposed ideas to enhance its generative ability. The results indicate that a modified training dynamic and a novel early stopping strategy improve the architecture’s capacity to synthesize data. The generated data presents realistic features with clear influences from its underlying dataset, and the inferred conclusions on subsequent analyses are similar to those based on the original data. Thus, the conclusion is that GANs has great potential to generate tabular data that can be considered a substitute for sensitive data, which could enable organizations to have more generous data sharing policies. / Med striktare förhållningsregler till hur data ska hanteras genom GDPR och PIPEDA har intresset för anonymiseringsmetoder för att censurera känslig data aktualliserats. En lovande teknik inom området återfinns i Generativa Motstridande Nätverk, en arkitektur som syftar till att generera data som återspeglar de statiska egenskaperna i dess underliggande dataset utan att äventyra datasubjektens integritet. Trots forskningsfältet unga ålder har man gjort stora framsteg i genereringsprocessen av så kallad syntetisk data, och numera finns det modeller som kan generera bilder av hög realistisk karaktär. Som ett steg framåt i forskningen har arkitekturen adopterats till nya domäner, och den här studien syftar till att undersöka dess förmåga att syntatisera finansiell tabelldata. I studien undersöks en framträdande modell inom forskningsfältet, CTGAN, tillsammans med två föreslagna idéer i syfte att förbättra dess generativa förmåga. Resultaten indikerar att en förändrad träningsdynamik och en ny optimeringsstrategi förbättrar arkitekturens förmåga att generera syntetisk data. Den genererade datan håller i sin tur hög kvalité med tydliga influenser från dess underliggande dataset, och resultat på efterföljande analyser mellan datakällorna är av jämförbar karaktär. Slutsatsen är således att GANs har stor potential att generera tabulär data som kan betrakatas som substitut till känslig data, vilket möjliggör för en mer frikostig delningspolitik av data inom organisationer.

This thesis addresses the issues of data sparsity in the sonar domain. A data pipeline is set up to generate and enhance sonar data. The possibilities and limitations of using cycleGAN as a tool to enhance simulated sonar images for the purpose of training neural networks for detection and classification is studied. A neural network is trained on the enhanced simulated sonar images and tested on real sonar images to evaluate the quality of these images.The novelty of this work lies in extending previous methods to a more general framework and showing that GAN enhanced simulations work for complex tasks on field data.Using real sonar images to enhance the simulated images, resulted in improved classification compared to a classifier trained on solely simulated images. / Denna rapport ämnar undersöka problemet med gles data för djupinlärning i sonardomänen. Ett dataflöde för att generera och höja kvalitén hos simulerad sonardata sätts upp i syfte att skapa en stor uppsättning data för att träna ett neuralt nätverk. Möjligheterna och begränsningarna med att använda cycleGAN för att höja kvalitén hos simulerad sonardata studeras och diskuteras. Ett neuralt nätverk för att upptäcka och klassificera objekt i sonarbilder tränas i syfte att evaluera den förbättrade simulerade sonardatan.Denna rapport bygger vidare på tidigare metoder genom att generalisera dessa och visa att metoden har potential även för komplexa uppgifter baserad på icke trivial data.Genom att träna ett nätverk för klassificering och detektion på simulerade sonarbilder som använder cycleGAN för att höja kvalitén, ökade klassificeringsresultaten markant jämfört med att träna på enbart simulerade bilder.

No description available.

In this study unintentionally doped GaN grown by hydride vapor phase epitaxy that exhibits a sharply delimited region of green color was investigated. Optical analysis was performed by absorption and photoluminescence spectroscopy. An absorption band between 1.5 and 2.0 eV was found to be responsible for the green color and was related to a sharp emission at 1.193 eV by luminescence and excitation spectroscopy. The appearance of both optical signatures in the region of green color was related to an increase of Cr contamination detected by secondary ion mass spectrometry. We propose that the origin of green color as well as the emission line at 1.193 eV is attributed to internal transitions of Cr⁴⁺.

Doctor of Philosophy / Department of Physics / Hongxing Jiang / III-nitride photonic devices such as photodetectors (PDs), light emitting diode (LEDs), solar cells and optical waveguide amplifiers were designed, fabricated and characterized. High quality AlN epilayers were grown on sapphire and n-SiC substrates by metal organic chemical vapor deposition and utilized as active DUV photonic materials for the demonstration of metal-semiconductor-metal (MSM) detectors, Schottky barrier detectors, and avalanche photodetectors (APDs). AlN DUV PDs exhibited peak responsivity at 200 nm with a very sharp cutoff wavelength at 207 nm and extremely low dark current (<10 fA), very high breakdown voltages, high responsivity, and more than four orders of DUV to UV/visible rejection ratio. AlN Schottky PDs grown on n-SiC substrates exhibited high zero bias responsivity and a thermal energy limited detectivity of about 1.0 x 1015 cm Hz1/2 W-1. The linear mode operation of AlN APDs with the shortest cutoff wavelength (210 nm) and a photocurrent multiplication of 1200 was demonstrated. A linear relationship between device size and breakdown field was observed for AlN APDs. Photovoltaic operation of InGaN solar cells in wavelengths longer than that of previous attainments was demonstrated by utilizing InxGa1−xN/GaN MQWs as the active layer. InxGa1-xN/GaN MQWs solar cells with x =0.3 exhibited open circuit voltage of about 2 V, a fill factor of about 60% and external quantum efficiency of 40% at 420 nm and 10% at 450 nm. The performance of InxGa1-xN/GaN MQWs solar cell was found to be highly correlated with the crystalline quality of the InxGa1-xN active layer. The possible causes of poorer PV characteristics for higher In content in InGaN active layer were explained. Photoluminescence excitation studies of GaN:Er and In0.06Ga0.94N:Er epilayers showed that Er emission intensity at 1.54 µm increases significantly as the excitation energy is tuned from below to above the energy bandgap of these epilayers. Current-injected 1.54 µm LEDs based on heterogeneous integration of Er-doped III-nitride epilayers with III-nitride UV LEDs were demonstrated. Optical waveguide amplifiers based on AlGaN/GaN:Er/AlGaN heterostructures was designed, fabricated, and characterized. The measured optical loss of the devices was ~3.5 cm−1 at 1.54 µm. A relative signal enhancement of about 8 dB/cm under the excitation of a broadband 365 nm nitride LED was achieved. The advantages and possible applications of 1.54 µm emitters and optical amplifiers based on Er doped III-nitrides in optical communications have been discussed.