Search

Bibliometric research is an important tool for examining the scientific output of various fields of study. By conducting such research, it is possible to see how the influences of different people, ideologies and discoveries have affected the scientific discourse. One way of doing this is through topic modelling, which consists of organizing the words that are used within a set of text data into different topics. To the knowledge of the author, no topic modelling study of Swedish philosophy had previously been conducted. For this reason, this study aimed to partially fill the gap by exploring the publications of one specific Swedish philosophical journal. Using Python, a topic model with 14 topics was created from the journal Filosofisk tidskrift. The change of these topics between the years 1980 and 2020 was examined. Specific attention was given to possible differences between analytic and Continental philosophy. To validate the results, an interview was also held with Fredrik Stjernberg, professor in theoretical philosophy. The results displayed a varied popularity and change for each topic. Too little Continental philosophy was discovered for a proper comparison, leading to the conclusion that Continental philosophy is not very influential in Swedish philosophical discourse. Future research should be conducted on peer-reviewed articles and be backed up by greater professional philosophical aid.

This article aims to present an experimental evaluation of an offline, geometry-aware aerial visual inspection framework, specifically in constrained environment, established for geometrically fractured objects, by employing an autonomous unmanned aerial vehicle (UAV), equipped with on-board sensors. Based on a model-centric approach, the proposed inspection framework, generates inspection viewpoints around the geometrically fractured object, subject to the augmented static bounds to prevent collisions. The novel framework of visual inspection, presented in this article, aims to mitigate challenges arising due to the spatially-constrained environment, such as limited configuration space and collision with the object under inspection, by accounting for the geometrical information of the vehicle to be inspected. The efficacy of the proposed scheme is experimentally evaluated through large scale field trials with a mining machine.

This article presents an overall system architecture for multi-robot coordination in a known environment. The proposed framework is structured around a task allocation mechanism that performs unlabeled multi-robot path assignment informed by 3D path planning, while using a nonlinear model predictive control(NMPC) for each unmanned aerial vehicle (UAV) to navigate along its assigned path. More specifically, at first a risk aware 3D path planner D∗+ is applied to calculate cost between each UAV agent and each target point. Then the cost matrix related to the computed trajectories to each goal is fed into the Hungarian Algorithm that solves the assignment problem and generates the minimum total cost. NMPC is implemented to control the UAV while satisfying path following and input constraints. We evaluate the proposed architecture in Gazebo simulation framework and the result indicates UAVs are capable of approaching their assigned target whilst avoiding collisions. ISBN för värdpublikation: 978-1-6654-0673-4 (electronic), 978-1-6654-0674-1 (print)

In this article, we propose a reactive constrained navigation scheme, with embedded obstacles avoidance for an unmanned aerial vehicle (UAV), for enabling navigation in obstacle-dense environments. The proposed navigation architecture is based on a nonlinear model predictive controller (NMPC) and utilizes an onboard 2-D LiDAR to detect obstacles and translate online the key geometric information of the environment into parametric constraints for the NMPC that constrain the available position space for the UAV. This article focuses also on the real-world implementation and experimental validation of the proposed reactive navigation scheme, and it is applied in multiple challenging laboratory experiments, where we also conduct comparisons with relevant methods of reactive obstacle avoidance. The solver utilized in the proposed approach is the optimization engine (OpEn) and the proximal averaged Newton for optimal control (PANOC) algorithm, where a penalty method is applied to properly consider obstacles and input constraints during the navigation task. The proposed novel scheme allows for fast solutions while using limited onboard computational power, which is a required feature for the overall closed-loop performance of a UAV and is applied in multiple real-time scenarios. The combination of built-in obstacle avoidance and real-time applicability makes the proposed reactive constrained navigation scheme an elegant framework for UAVs that is able to perform fast nonlinear control, local path planning, and obstacle avoidance, all embedded in the control layer.

This compilation thesis presents an overarching framework on the utilization of nonlinear model predictive control(NMPC) for various applications in the context of Unmanned Aerial Vehicle (UAV) path planning and collision avoidance. Fast and novel optimization algorithms allow for NMPC formulations with high runtime requirement, as those posed by controlling UAVs, to also have sufficiently large prediction horizons as to in an efficient manner integrate collision avoidance in the form of set-exclusion constraints that constrain the available position-space of the robot. This allows for an elegant merging of set-point reference tracking with the collision avoidance problem, all integrated in the control layer of the UAV. The works included in this thesis presents the UAV modeling, cost functions, constraint definitions, as well as the utilized optimization framework. Additional contributions include the use case on multi-agent systems, how to classify and predict trajectories of moving (dynamic) obstacles, as well as obstacle prioritization when an aerial agent is in the precense of more obstacles, or other aerial agents, than can reasonably be defined in the NMPC formulation. For the cases of dynamic obstacles and for multi-agent distributed collision avoidance this thesis offers extensive experimental validation of the overall NMPC framework. These works push the limits of the State-of-the-Art regarding real-time real-life implementations of NMPC-based collision avoidance. The works also include a novel RRT-based exploration framework that combines path planning with exploration behavior. Here, a multi-path RRT * planner plans paths to multiple pseudo-random goals based on a sensor model and evaluates them based on the potential information gain, distance travelled, and the optimimal actuation along the paths.The actuation is solved for as as the solutions to a NMPC problem, implying that the nonlinear actuator-based and dynamically constrained UAV model is considered as part of the combined exploration plus path planning problem. To the authors best knowledge, this is the first time the optimal actuation has been considered in such a planning problem. For all of these applications, the utilized optimization framework is the Optimization Engine: a code-generation framework that generates a custom Rust-based solver from a specified model, cost function, and constraints. The Optimization Engine solves general nonlinear and nonconvex optimization problems, and in this thesis we offer extensive experimental validation of the utilized Proximal-Averaged Newton-type method for Optimal Control (PANOC) algorithm as well as both the integrated Penalty Method and Augmented Lagrangian Method for handling the nonlinear nonconvex constraints that result from collision avoidance problems.

We investigate the modelling of polyphonic “tracker music” using deep neural networks. Tracker music is a music storage format invented in the late 1980s for use on that time’s home computers and is often used for storing synthesized electronic music. Tracker music differs significantly from other music formats and has properties that makes it both harder and easier to use for training neural networks than other formats. This makes it interesting to explore what methods are most suitable for extracting musical information from the format. As far as we know, we are the first to explore how to use tracker music for music generation. We design a method for turning tracker music into sequential data usable for training neural networks. The sequential nature of the data means that musically unaware sequence models can be used for training. The method is general and can be applied to other kinds of symbolic music. We then compile a dataset of about 20 000 freely available instrumental songs in the tracker format MOD, downloaded from the website The Mod Archive. We use the dataset to train several different sequence models, including a Long Short-Term Memory (LSTM) network and a Transformer model. We evaluate the models using a sequence completion task and we investigate the statistical properties of the output. We also conduct a listener study involving some 100 participants to determine how often music generated by the models is preferred over human-composed music. The listener study’s result indicates that music generated by the models trained on the dataset is sometimes competitive with music composed by humans. We conclude that neural networks for music generation can be trained using tracker music using our proposed conversion method, but that it is cumbersone. Due to how the tracker music format is constructed it is significantly more difficult to get musical information out of it than we initially thought. Vi undersöker hur man bäst använder sig av trackermusik för att träna neurala nätverk till att generera polyfonisk musik. Trackermusik kan sägas både vara en speciell instrumental musikgenre och ett speciellt musikformat. Formatet som skiljer sig markant från exempelvis MIDI och MP3 har en del egenskaper som gör det svårare och andra som gör det lättare att träna neurala nätverk med det än med jämförbara musikformat. Därför är det intressant att utforska vilka metoder som är bäst att använda för att utvinna musikinformation ur formatet. Så vitt vi vet har ämnet inte utforskats tidigare och vårt utforskande av det är vår uppsats centrala bidrag till forskningen kring musikgenerering med neurala nätverk. I uppsatsen föreslår vi en metod som konverterar trackermusik till ett sekventiellt format som är lämpligt att använda för att träna neurala nätverk. Vi demonstrerar också att metoden fungerar i praktiken genom att träna ett antal neurala nätverk med en samling på cirka 20 000 instrumentala sånger i trackermusiklagringsformatet MOD som vi sedan utvärderar. Utvärderingen består bland annat i en lyssnarstudie. Resultatet av den visar att den musik som genereras av tre neurala nätverk som vi tränar med trackermusiksamlingen ibland föredras av lyssnare över musik skapad av människor. Vi drar slutsatsen att musikgenererande neurala nätverk kan tränas med hjälp av trackermusik med vår föreslagna konverteringsmetod, men att det är bökigt. På grund av hur trackermusik är uppbyggt och organiserat är det betydligt svårare än vad vi inledningsvis trodde att utvinna musikinformation ur det.

Using robots to navigate through un-mapped environments, specially man-made infrastructures, for the purpose of exploration or inspection is a topic that has gathered a lot of interest in the last years. Micro Aerial Vehicles (MAV's) have the mobility and agility to move quickly and access hard-to-reach areas where ground robots would fail, but using MAV's for that purpose comes with its own set of problems since any collision with the environment results in a crash. The control architecture used in a MAV for such a task needs to perform obstacle avoidance and on-line path-planning in an unknown environment with low computation times as to not lose stability. In this thesis a Non-linear Model Predictive Controller (NMPC) for obstacle avoidance and path-planning on an aerial platform will be established. Included are methods for constraining the available state-space, simulations of various obstacle avoidance scenarios for single and multiple MAVs and experimental validation of the proposed control architecture. The validity of the proposed approach is demonstrated through multiple experimental and simulation results. In these approaches, the positioning information of the obstacles and the MAV are provided by a motion-capture system. The thesis will conclude with the demonstration of an experimental validation of a centralized NMPC for collision avoidance of two MAV's.

Great strides have been made in the state of the art performance of musicial onset detection in recent years with better and better detectors being invented at a fast pace. The current top spot is held by Schlüter and Böck, who in 2014 presented a detector based on a convolutional neural network (CNN) that attained an F-score of 90.3% (Precision 91.7%, 88.9% recall) on a commonly used dataset [1]. In 2018 two researchers, Gong and Serra, tried to replicate their result but only reached an F-score of 86.67% (precision and recall values weren’t reported) [2], a significantly worse result than Schlüter and Böck’s. In comparison a 2013 detector based on a recurrent network, also designed by Schlüter and Böck, achieved an F-score of 87.3% [3]. Gong and Serra’s result casts doubt on the 90.3% figure reported by Schlüter and Böck. We therefore try to shed some light on the question of what the state of the art performance in musical onset detection is by posing and answering the question; can Schlüter and Böck’s result be reproduced? Our answer is “Maybe – but we were unable to!” which is perhaps the only result possible since you can’t prove a negative. We trained the CNN architecture three times and obtained F-scores of 85.0%, 85.8% and 85.6%. For the RNN architecture, which we also tried to reproduce, we obtained the scores 86.3%, 86.3% and 86.3%. Due to omission of details that perhaps were significant from the referenced articles, we weren’t able to recreate Schlüter and Böck’s architectures exactly and had to make some “educated guesses.” It is possible that those guesses caused performance to suffer. Nevertheless, we believe that our work is worthwhile because it demonstrates how infuriatingly difficult it is in deep learning for researchers to reproduce each others work. Ansatsdetektion är ett område inom musikanalysen som går ut på att bestämma när händelser i ljuddata inträffar. Området har utvecklats betydligt på sistone och flera nya metoder har föreslagits. Dessa har gradvis förbättrat den bästa detekteringsförmågan. Nuvarande rekord innehas av Schlüter och Böck som år 2014 presenterade en detektions- metod based på ett faltningsnätverk (Convolutional Neural Network, CNN) som uppnådde ett F-värde på 90,3 % (med precision 91,7 % och täckning 88,9 %) på en ofta använd mängd data [1]. Men år 2018 misslyckades ett annat forskarpar, Gong och Serra, med att replikera de förstnämndas resultat. Deras F-värde stannade på 86,67 % (de rapporterade varken precision eller täckning) [2] vilket är ett signifikant sämre resultat än Schlüter och Böcks. Som jämförelse kan nämnas att år 2013 presenterade Schlüter och Böck en detektionsmetod based på ett återkommande neuralt nätverk (Recurrent Neural Network, RNN) som hade F-värdet 87,3 % [3]. Gong och Serras resultat gör att Schlüter och Böcks 90,3 %-resultat kan ifrågasättas. I den här avhandlingen försöker vi därför att ta reda på om de sistnämndas resultat håller. Detta gör vi genom att implementera de nätverk som föreslagits och se om vi kan få lika bra detektionsförmåga på de datamängder som Schlüter och Böck själva använt sig av. Kan deras resultat replikeras? Vårt svar är “Kanske – men vi kunde det inte!” Möjligtvis är det det enda säkra som kan sägas eftersom vi misslyckades med att replikera deras resultat. De tre faltningsnätverk vi tränade fick F-värdena 85,0%, 85,8 %, och 85,6 %, cirka fem procentenheter lägre än Schlüter och Böcks 90,3 %. Vi försökte också replikera deras resultat för det återkommande neurala nätverket och där fick vi värdena 86,3 %, 86,3% och 86,3 %. Även dessa värden är sämre än Schlüter och Böcks, men här är skillnaden bara en procentenhet. Dock saknades detaljinformation i de korta artiklar författade av Schlüter och Böck och Böck et al. som vi använde oss av för att förstå metoderna. Därför var vi tvungna att gissa oss till vissa detaljer såsom parametrar för inlärningshastighet osv och vi kan därför inte garantera att de implementationer vi utvärderat är exakt desamma som författarnas. Detta kan vara en delförklaring till vårt sämre resultat. Vi menar ändå att vårt arbete är värdefullt eftersom det visar hur otroligt svårt det är att replikera resultat inom området djupinlärning.