The last decade has seen great advances within the field of artificial intelligence. One of the most noteworthy areas is that of deep learning, which is nowadays used in everything from self driving cars to automated cancer screening. During the same time, the amount of spatial data encompassing not only two but three dimensions has also grown and whole cities and countries are being scanned. Combining these two technological advances enables the creation of detailed maps with a multitude of applications, civilian as well as military.This thesis aims at combining two data sources covering most of Sweden; laser data from LiDAR scans and surface model from aerial images, with deep learning to create maps of the terrain. The target is to learn a simplified version of orienteering maps as these are created with high precision by experienced map makers, and are a representation of how easy or hard it would be to traverse a given area on foot. The performance on different types of terrain are measured and it is found that open land and larger bodies of water is identified at a high rate, while trails are hard to recognize.It is further researched how the different densities found in the source data affect the performance of the models, and found that some terrain types, trails for instance, benefit from higher density data, Other features of the terrain, like roads and buildings are predicted with higher accuracy by lower density data.Finally, the certainty of the predictions is discussed and visualised by measuring the average entropy of predictions in an area. These visualisations highlight that although the predictions are far from perfect, the models are more certain about their predictions when they are correct than when they are not.

The population of northern pike in the Baltic sea has seen a drasticdecrease in numbers in the last couple of decades. The reasons for this are believed to be many, but the majority of them are most likely anthropogenic. Today, many measures are being taken to prevent further decline of pike populations, ranging from nutrient runoff control to habitat restoration. This inevitably gives rise to the problem addressed in this project, namely: how can we best monitor pike populations so that it is possible to accurately assess and verify the effects of these measures over the coming decades? Pike is currently monitored in Sweden by employing expensive and ineffective manual methods of individual marking of pike by a handful of experts. This project provides evidence that such methods could be replaced by a Convolutional Neural Network (CNN), an automatic artificial intelligence system, which can be taught how to identify pike individuals based on their unique patterns. A neural net simulates the functions of neurons in the human brain, which allows it to perform a range of tasks, while a CNN is a neural net specialized for this type of visual recognition task. The results show that the CNN trained in this project can identify pike individuals in the provided data set with upwards of 90% accuracy, with much potential for improvement.

No description available.

Computer vision object detection is the task of detecting and identifying objects present in an image or a video sequence. Models based on artificial convolutional neural networks are commonly used as detector models. Object detection precision and inference efficiency are crucial for surveillance-based applications. A decrease in the detector model complexity as well as in the complexity of the post-processing computations promotes increased inference efficiency. Modern object detectors for surveillance applications usually make use of a regression algorithm and bounding box priors referred to as anchor boxes to compute bounding box proposals, and the proposal selection algorithm contributes to the computational cost at inference. In this study, an anchor-free and low complexity deep learning detector model was implemented within a surveillance applications setting, and was evaluated and compared to a reference baseline state-of-the-art anchor-based object detector. A key-point-based detector model (CenterNet), predicting Gaussian distribution based object centers, was selected for the evaluation against the baseline. The surveillance applications adapted anchor-free detector exhibited a factor 2.4 lower complexity than the baseline detector. Further, a significant redistribution to shorter post-processing times was demonstrated at inference for the anchor-free surveillance adapted CenterNet detector, giving a modal values factor 0.6 of the baseline detector post-processing time. Furthermore, the surveillance adapted CenterNet model was shown to outperform the baseline in terms of detection precision for several surveillance applications relevant classes and for objects of smaller spatial scale.

Brain tumor is a disease characterized by uncontrolled growth of abnormal cells inthe brain. The brain is responsible for regulating the functions of all other organs,hence, any atypical growth of cells in the brain can have severe implications for itsfunctions. The number of global mortality in 2020 led by cancerous brains was estimatedat 251,329. However, early detection of brain cancer is critical for prompttreatment and improving patient’s quality of life as well as survival rates. Manualmedical image classification in diagnosing diseases has been shown to be extremelytime-consuming and labor-intensive. Convolutional Neural Networks (CNNs) hasproven to be a leading algorithm in image classification outperforming humans. Thispaper compares five CNN architectures namely: VGG-16, VGG-19, AlexNet, EffecientNetB7,and ResNet-50 in terms of performance and accuracy using transferlearning. In addition, the authors discussed in this paper the economic impact ofCNN, as an AI approach, on the healthcare sector. The models’ performance isdemonstrated using functions for loss and accuracy rates as well as using the confusionmatrix. The conducted experiment resulted in VGG-19 achieving best performancewith 97% accuracy, while EffecientNetB7 achieved worst performance with93% accuracy. / Hjärntumör är en sjukdom som kännetecknas av okontrollerad tillväxt av onormalaceller i hjärnan. Hjärnan är ansvarig för att styra funktionerna hos alla andra organ,därför kan all onormala tillväxt av celler i hjärnan ha allvarliga konsekvenser för dessfunktioner. Antalet globala dödligheten ledda av hjärncancer har uppskattats till251329 under 2020. Tidig upptäckt av hjärncancer är dock avgörande för snabb behandlingoch för att förbättra patienternas livskvalitet och överlevnadssannolikhet.Manuell medicinsk bildklassificering vid diagnostisering av sjukdomar har visat sigvara extremt tidskrävande och arbetskrävande. Convolutional Neural Network(CNN) är en ledande algoritm för bildklassificering som har överträffat människor.Denna studie jämför fem CNN-arkitekturer, nämligen VGG-16, VGG-19, AlexNet,EffecientNetB7, och ResNet-50 i form av prestanda och noggrannhet. Dessutom diskuterarförfattarna i studien CNN:s ekonomiska inverkan på sjukvårdssektorn. Modellensprestanda demonstrerades med hjälp av funktioner om förlust och noggrannhetsvärden samt med hjälp av en Confusion matris. Resultatet av det utfördaexperimentet har visat att VGG-19 har uppnått bästa prestanda med 97% noggrannhet,medan EffecientNetB7 har uppnått värsta prestanda med 93% noggrannhet.

Hypothesis testing has long been a formal and standardized process. Hypothesis generation, on the other hand, remains largely informal. This thesis assess whether eXplainable AI (XAI) can aid in the standardization of hypothesis generation through its utilization as a hypothesis generating tool for medical research. We produce XAI heat maps for a Convolutional Neural Network (CNN) trained to classify Microsatellite Instability (MSI) in colon and gastric cancer with four different XAI methods: Guided Backpropagation, VarGrad, Grad-CAM and Sobol Attribution. We then compare these heat maps with pathology annotations in order to look for differences to turn into new hypotheses. Our CNN successfully generates non-random XAI heat maps whilst achieving a validation accuracy of 85% and a validation AUC of 93% – as compared to others who achieve a AUC of 87%. Our results conclude that Guided Backpropagation and VarGrad are better at explaining high-level image features whereas Grad-CAM and Sobol Attribution are better at explaining low-level ones. This makes the two groups of XAI methods good complements to each other. Images of Microsatellite Insta- bility (MSI) with high differentiation are more difficult to analyse regardless of which XAI is used, probably due to exhibiting less regularity. Regardless of this drawback, our assessment is that XAI can be used as a useful hypotheses generating tool for research in medicine. Our results indicate that our CNN utilizes the same features as our basic pathology annotations when classifying MSI – with some additional features of basic pathology missing – features which we successfully are able to generate new hypotheses with.

Självkörande fordon kan minska trafikstockningar och minska antalet trafikrelaterade olyckor. Då det i framtiden kommer att finnas miljontals autonoma fordon krävs en bättre förståelse av omgivningen. Syftet med detta projekt är att skapa ett externt automatiskt trafikledningssystem som kan upptäcka och spåra 3D-objekt i en komplex trafiksituation för att senare skicka beteendet från dessa objekt till ett större projekt som hanterar med att 3D-modellera trafiksituationen. Projektet använder sig av Tensorflow ramverket och YOLOv3 algoritmen. Projektet använder sig även av en kamera för att spela in trafiksituationer och en dator med Linux som operativsystem. Med hjälp av metoder som vanligen används för att skapa ett automatiserat trafikledningssystem utvärderades ett målföljningssystem. De slutliga resultaten visar att systemet är relativt instabilt och ibland inte kan känna igen vissa objekt. Om fler bilder används för träningsprocessen kan ett robustare och mycket mer tillförlitligt system utvecklas med liknande metodik. / Autonomous vehicles can decrease traffic congestion and reduce the amount of traffic related accidents. As there will be millions of autonomous vehicles in the future, a better understanding of the environment will be required. This project aims to create an external automated traffic system that can detect and track 3D objects within a complex traffic situation to later send these objects’ behavior for a larger-scale project that manages to 3D model the traffic situation. The project utilizes Tensorflow framework and YOLOv3 algorithm. The project also utilizes a camera to record traffic situations and a Linux operated computer. Using methods commonly used to create an automated traffic management system was evaluated. The final results show that the system is relatively unstable and can sometimes fail to recognize certain objects. If more images are used for the training process, a more robust and much more reliable system could be developed using a similar methodology.

The research uses cameras mounted to a vehicle to capture geotagged images while conducting a transect survey. The images from two capture dates were manually classified into different classes of previous crop, tillage systems, residue cover, and cover crop utilization. The raw data was compared against the Indiana Cropland Transect Survey and the USDA-NASS Cropland Data Layer. The symmetric Kullback-Liebler divergence method was used to compared the distributions looking for similarities. <div><br></div><div>The manually classified data was then used to build satellite segmentation models using artificial neural networks , decision trees, k nearest neighbors, random forests, and support vector machine methods. The models were compared using overall accuracy, kappa coefficient, specificity, sensitivity, positive prediction value, and negative prediction value. The best model for each category of previous crop, tillage system, residue cover, and cover crop was used to segment a Sentenial-2 imagery downloaded from Copernicus Open Access hub. The results of the segment were compared by looking at the agreement at individual pixel locations from the segmented raster to the manually classified data and the Indiana Cropland Transect Survey. </div><div><br></div><div>Finally, all the images captured were used to being the development of a automated image classifier using nested convolutional neural networks (CNN). A small set of images was used to build the CNN. That model when then make prediction on new unclassified images. The predictions were manually checked. The check images were used to the to build the training and validation pools for the models. The first network divided the images into field or not field.</div><div>The second branch was field images divided in to images containing green growing plants of brown dead plants or residues. The final branch was determining the amount of surface cover left on a field. The results from each run of the training process were saved and used to assess model performance looking at accuracy and loss.</div>

This study aims to look at whether or not it is sufficient to only use smartphone sensors to judge if someone who is driving a car is driving aggressively or smoothly. To determine this, data were first collected from the accelerometer, gyroscope, magnetometer and GPS sensors in the smartphone as well as values based on these sensors from the iOS operating system. After this the data, together with synthesized data based on the collected data, were used to train an artificial neural network.The results indicate that it is possible to give a binary judgment on aggressive or smooth driving with a 97% accuracy, with little model overfitting. The conclusion of this study is that it is sufficient to only use smartphone sensors to make a judgment on the drive. / Den här studien ämnar till att bedöma huruvida smartphonesensorer är tillräckliga för att avgöra om någon kör en bil aggressivt eller mjukt. För att kunna avgöra detta så samlades först data in från accelerometer, gyroskop, magnetometer och GPS-sensorerna i en smartphone, tillsammans med värden baserade på dessa data från iOS-operativ-systemet. Efter den datan var insamlad tränades ett artificiellt neuronnät med datan.Resultaten indikerar att det är möjligt att ge ett binärt utlåtande om aggressiv kontra mjuk körning med 97% säkerhet, och med liten överanpassning. Detta innebär att det är tillräckligt att enbart använda smartphonesensorer för att avgörande om körningen var mjuk eller aggressiv.

In this paper an analysis is conducted regarding whether a higher classification accuracy of facial expressions are possible. The approach used is that the seven basic emotional states are combined into a binary classification problem. Five different machine learning algorithms are implemented: Support vector machines, Extreme learning Machine and three different Convolutional Neural Networks (CNN). The utilized CNN:S were one conventional, one based on VGG16 and transfer learning and one based on residual theory known as RESNET50. The experiment was conducted on two datasets, one small containing no contamination called JAFFE and one big containing contamination called FER2013. The highest accuracy was achieved with the CNN:s where RESNET50 had the highest classification accuracy. When comparing the classification accuracy with the state of the art accuracy an improvement of around 0.09 was achieved on the FER2013 dataset. This dataset does however include some ambiguities regarding what facial expression is shown. It would henceforth be of interest to conduct an experiment where humans classify the facial expressions in the dataset in order to achieve a benchmark.