Nanopore sequencing, a recently developed methodfor DNA sequencing, involves applying a constant electricfield over a membrane and translocating single-stranded DNAmolecules through membrane pores. This results in an electricalsignal, which is dependent on the structure of the DNA. The aimof this project is to train and evaluate a non-causal temporalconvolution neural network in order to accurately translate suchelectrical raw signal into the corresponding nucleotide sequence.The training dataset is sampled from the E. coli bacterial genomeand the phage Lambda virus. We implemented and evaluatedseveral different temporal convolutional architectures. Using anetwork with five residual blocks with five convolutional layersin each block yields maximum performance, with a predictionaccuracy of 76.1% on unseen test data. This result indicates thata temporal convolution network could be an effective way tosequence DNA data. / Nanopore sequencing är en nyligen utvecklad metod för DNA-sekvensering som innebär att man applicerar ett konstant elektriskt fält över ett membran och translokerar enkelsträngade DNA-molekyler genom membranporer. Detta resulterar i en elektrisk signal som beror på DNA-strukturen.  Målet med detta projekt är att träna och evaluera icke-kausula ”temporal convolutional networks” som ska kunna översätta denna ofiltrerade elektriska signalen till den motsvarande nukleotidsekvensen. Träningsdatan är ett urval av genomen från bakterien E. coli och viruset phage Lambda. Vi implementerade och utvärderade ett antal olika nätverksstrukturer. Ett nätverk med fem residuala block med fem faltande lager i varje block gav maximal prestation, med en precision på 76.1% på testdata. Detta resultat indikerar att ett ”temporal convolution network” skulle kunna vara ett effektivt sätt att sekvensera DNA. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

This work uses a Deep Learning architecture, specifically a state-of-the-art Temporal Convolutional Network, to track the beat and downbeat positions in jazz-ensemble recordings to derive their metrical grid. This network architecture has been used successfully for general beat tracking purposes. However, the jazz genre presents difficulties for this Music Information Retrieval sub-task due to its inherent complexity, and there is a lack of dedicated sets for evaluating a model’s beat tracking performance for different playstyles of this specific music genre. We present a methodology in which we trained a PyTorch implementation of the original architecture with a recalculated binary cross-entropy loss that helps boost the model’s performance compared to a standard trained version. In addition, we retrained these two models using source-separated drums and bass tracks from jazz recordings to improve performance. We further improved the model’s performance by calibrating rhythm parameters using a priori knowledge that narrows the model’s prediction range. Finally, we proposed a novel jazz dataset comprised of recordings from the same jazz piece played with different styles and used this to evaluate the performance of this methodology. We also evaluate a novel sample with tempo variations to demonstrate the architecture’s versatility. This methodology, or parts of it, can be exported to other research work and music information tools that perform beat tracking or other similar Music Information Retrieval sub-tasks. / Vi använde en Deep Learning-arkitektur för att spåra beat- och downbeatpositionerna i jazz-ensembleinspelningar för att härleda deras metriska rutnät. Denna nätverksarkitektur har använts framgångsrikt för allmän taktspårning. Men jazzgenren uppvisar svårigheter för denna deluppgift för återhämtning av musikinformation på grund av dess inneboende komplexitet, och det finns en brist på dedikerade datauppsättningar för att utvärdera en modells prestanda för olika spelstilar av denna specifika musikgenre. Vi presenterar en metod där vi tränade modellen med en omräknad binär korsentropiförlust som hjälper till att öka modellens prestanda jämfört med en utbildad standardversion. Dessutom tränade vi om dessa två modeller med hjälp av källseparerade spår från jazzinspelningar för att förbättra resultaten. Vi förbättrade modellens prestanda ytterligare genom att kalibrera parametrar med hjälp av a priori kunskap. Slutligen föreslog vi en ny jazzdatauppsättning bestående av inspelningar från samma jazzstycke som spelades med olika stilar och använde detta för att utvärdera hur denna metod fungerar. Vi utvärderar också ett nytt prov med tempovariationer för att visa arkitekturens mångsidighet. Denna metodik, eller delar av den, kan exporteras till andra forskningsarbeten och musikinformationsverktyg som utför beat tracking eller andra liknande Music Information Retrieval underuppgifter.

Recent studies show that deep learning can be employed to learn from sensor data to improve accuracy and robustness of sensor fusion algorithms. In the same vein, in this thesis we use a state-of-the-art temporal convolution network to predict zero velocity updates (ZUPT) from raw inertial measurement unit (IMU) signals, and use the network output to improve the performance of an optimization-based pose estimator. Experiments were conducted on publicly available datasets, and results show that (i) the network can distinguish a car in motion vs. a car standing still by observing an IMU signal, and (ii) that ZUPT detection enhances the observability of states in the optimization-based pose estimation, thus reducing local drift. / Nyligen gjorda studier visar att djupinlärning kan användas för att lära av sensordata för att förbättra noggrannhet och robusthet hos sensorfusionsalgoritmer. På samma sätt använder vi i denna avhandling en tidsberoende faltnings neuronnätsmodell (TCN) för att detektera om ett fordon står stilla även kallat zero velocity updates (ZUPT) från IMU rå- data och använder neuronnätsprediktionen för att förbättra prestandan hos en optimeringsbaserad positionsuppskattning. Experiment utfördes på allmänt publicerade datamängder, och resultaten visar att (i) neuronnätsmodellen kan läras till att urskilja en bil i rörelse kontra en bil som står stilla genom att observera en IMU- signal, och (ii) att ZUPT- detektering förbättrar observerbarheten för tillstånd i den optimeringsbaserade positioneringsuppskattningen, vilket minskar lokal drift.

No description available.

The prediction of polypharmacy side effects is crucial to reduce the mortality and morbidity of patients suffering from complex diseases. However, its experimental prediction is unfeasible due to the many possible drug combinations, leaving in silico tools as the most promising way of addressing this problem. This thesis improves the performance and robustness of a state-of-the-art graph convolutional network designed to predict polypharmacy side effects, by feeding it with complexity properties of the drug-protein network. The modifications also involve the creation of a direct pipeline to reproduce the results and test it with different datasets. / För att minska dödligheten och sjukligheten hos patienter som lider av komplexa sjukdomar är det avgörande att kunna förutsäga biverkningar från polyfarmaci. Att experimentellt förutsäga biverkningarna är dock ogenomförbart på grund av det stora antalet möjliga läkemedelskombinationer, vilket lämnar in silico-verktyg som det mest lovande sättet att lösa detta problem. Detta arbete förbättrar prestandan och robustheten av ett av det senaste grafiska faltningsnätverken som är utformat för att förutsäga biverkningar från polyfarmaci, genom att mata det med läkemedel-protein-nätverkets komplexitetsegenskaper. Ändringarna involverar också skapandet av en direkt pipeline för att återge resultaten och testa den med olika dataset.

Nanopore DNA sequencing is a novel method forsequencing DNA where an electronic signal is modulated bynucleotides passing through nanosized pores embedded in a mem-brane. While current state-of-the-art solutions employ recurrentneural networks to analyse the signal, temporal convolutionalnetworks have recently been shown to match or outperformrecurrent networks in signal processing tasks. In this project, weinvestigate the performance of temporal convolutional networkson a simplified version of the sequencing task, where thegoal is to predict the nucleotides passing through the pore ateach instance in time, without reconstructing the correspondingDNA sequence. The impact of several network parameters onpredictive performance is analysed to determine an optimalarchitecture. While the implemented networks are shown tobe proficient at predicting nucleotides within the pore, thecurrent implementation is unlikely to outperform state-of-the-art solutions without further improvement. / En nyligen utvecklad metod för att sekvensera DNA innefattar att en elektrisk signal moduleras genom att nukleotider passerar genom porer i nanostorlek. I kommersiella lösningar analyseras denna signal med hjälp av maskininlärning via Recurrent Neural Networks, men en variant av neruala nätverk som kallas Temporal Convolution Networks har nyligen har visat sig ha bättre prestanda jämfört med Recurrent Networks för olika typer av signalbehandlingsproblem. Målet med detta projekt är att undersöka användbarheten av Temporal Convolutional Networks för en förenklad version av DNA-sekvensering, där uppdraget endast är att identifera de nukleotider som passerar genom poren vid varje given tidpunkt, istället för att rekonstruera en komplett DNA-sekvens. För att kunna bestämma en optimal arkitektur på nätverket så undersöks effekten av flera olika parametrar. De implementerade nätverken visas ha god förmåga att klassificera nukleotider, men är troligtvis i behov av ytterligare förbättringar för att kunna konkurrera med nuvarande kommersiella lösningar. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

Generation of realistic skydome images We aim to generate realistic images of the sky with clouds using generative adversarial networks (GANs). We explore two GAN architectures, ProGAN and StyleGAN, and find that StyleGAN produces significantly better results. We also propose a novel architecture SuperGAN which aims to generate images at very high resolutions, which cannot be efficiently handled using state-of-art architectures. 1

No description available.

Nowadays, we have millions of cameras in public places such as traffic junctions, railway stations etc., and capturing video data round the clock. This humongous data has resulted in an increased need for automation of visual surveillance. Analysis of crowd and traffic flows is an important step towards achieving this goal. In this work, we present our algorithms for identifying and segmenting dominant ows in surveillance scenarios. In the second part, we present our work aiming at predicting the visual saliency. The ability of humans to discriminate and selectively pay attention to few regions in the scene over the others is a key attentional mechanism. Here, we present our algorithms for predicting human eye fixations and segmenting salient objects in the scene. (i) Flow Analysis in Surveillance Videos: We propose algorithms for segmenting flows of static and dynamic nature in surveillance videos in an unsupervised manner. In static flows scenarios, we assume the motion patterns to be consistent over the entire duration of video and analyze them in the compressed domain using H.264 motion vectors. Our approach is based on modeling the motion vector field as a Conditional Random Field (CRF) and obtaining oriented motion segments which are merged to obtain the final flow segments. This approach in compressed domain is shown to be both accurate and computationally efficient. In the case of dynamic flow videos (e.g. flows at a traffic junction), we propose a method for segmenting the individual object flows over long durations. This long-term flow segmentation is achieved in the framework of CRF using local color and motion features. We propose a Dynamic Time Warping (DTW) based distance measure between flow segments for clustering them and generate representative dominant ow models. Using these dominant flow models, we perform path prediction for the vehicles entering the camera's field-of-view and detect anomalous motions. (ii) Visual Saliency Prediction using Deep Convolutional Neural Networks: We propose a deep fully convolutional neural network (CNN) - DeepFix, for accurately predicting eye fixations in the form of saliency maps. Unlike classical works which characterize the saliency map using various hand-crafted features, our model automatically learns features in a hierarchical fashion and predicts saliency map in an end-to-end manner. DeepFix is designed to capture visual semantics at multiple scales while taking global context into account. Generally, fully convolutional nets are spatially invariant which prevents them from modeling location dependent patterns (e.g. centre-bias). Our network overcomes this limitation by incorporating a novel Location Biased Convolutional layer. We experimentally show that our network outperforms other recent approaches by a significant margin. In general, human eye fixations correlate with locations of salient objects in the scene. However, only a handful of approaches have attempted to simultaneously address these related aspects of eye fixations and object saliency. In our work, we also propose a deep convolutional network capable of simultaneously predicting eye fixations and segmenting salient objects in a unified framework. We design the initial network layers, shared between both the tasks, such that they capture the global contextual aspects of saliency, while the deeper layers of the network address task specific aspects. Our network shows a significant improvement over the current state-of-the-art for both eye fixation prediction and salient object segmentation across a number of challenging datasets.

Le discours haineux est un problème sérieux sur les média sociaux. Dans ce mémoire, nous étudions le problème de détection automatique du langage haineux sur réseaux sociaux. Nous traitons ce problème comme un problème de classification de textes. La classification de textes a fait un grand progrès ces dernières années grâce aux techniques d’apprentissage profond. En particulier, les modèles utilisant un mécanisme d’attention tel que BERT se sont révélés capables de capturer les informations contextuelles contenues dans une phrase ou un texte. Cependant, leur capacité à saisir l’information globale sur le vocabulaire d’une langue dans une application spécifique est plus limitée. Récemment, un nouveau type de réseau de neurones, appelé Graph Convolutional Network (GCN), émerge. Il intègre les informations des voisins en manipulant un graphique global pour prendre en compte les informations globales, et il a obtenu de bons résultats dans de nombreuses tâches, y compris la classification de textes. Par conséquent, notre motivation dans ce mémoire est de concevoir une méthode qui peut combiner à la fois les avantages du modèle BERT, qui excelle en capturant des informations locales, et le modèle GCN, qui fournit les informations globale du langage. Néanmoins, le GCN traditionnel est un modèle d'apprentissage transductif, qui effectue une opération convolutionnelle sur un graphe composé d'éléments à traiter dans les tâches (c'est-à-dire un graphe de documents) et ne peut pas être appliqué à un nouveau document qui ne fait pas partie du graphe pendant l'entraînement. Dans ce mémoire, nous proposons d'abord un nouveau modèle GCN de vocabulaire (VGCN), qui transforme la convolution au niveau du document du modèle GCN traditionnel en convolution au niveau du mot en utilisant les co-occurrences de mots. En ce faisant, nous transformons le mode d'apprentissage transductif en mode inductif, qui peut être appliqué à un nouveau document. Ensuite, nous proposons le modèle Interactive-VGCN-BERT qui combine notre modèle VGCN avec BERT. Dans ce modèle, les informations locales captées par BERT sont combinées avec les informations globales captées par VGCN. De plus, les informations locales et les informations globales interagissent à travers différentes couches de BERT, ce qui leur permet d'influencer mutuellement et de construire ensemble une représentation finale pour la classification. Via ces interactions, les informations de langue globales peuvent aider à distinguer des mots ambigus ou à comprendre des expressions peu claires, améliorant ainsi les performances des tâches de classification de textes. Pour évaluer l'efficacité de notre modèle Interactive-VGCN-BERT, nous menons des expériences sur plusieurs ensembles de données de différents types -- non seulement sur le langage haineux, mais aussi sur la détection de grammaticalité et les commentaires sur les films. Les résultats expérimentaux montrent que le modèle Interactive-VGCN-BERT surpasse tous les autres modèles tels que Vanilla-VGCN-BERT, BERT, Bi-LSTM, MLP, GCN et ainsi de suite. En particulier, nous observons que VGCN peut effectivement fournir des informations utiles pour aider à comprendre un texte haiteux implicit quand il est intégré avec BERT, ce qui vérifie notre intuition au début de cette étude. / Hate speech is a serious problem on social media. In this thesis, we investigate the problem of automatic detection of hate speech on social media. We cast it as a text classification problem. With the development of deep learning, text classification has made great progress in recent years. In particular, models using attention mechanism such as BERT have shown great capability of capturing the local contextual information within a sentence or document. Although local connections between words in the sentence can be captured, their ability of capturing certain application-dependent global information and long-range semantic dependency is limited. Recently, a new type of neural network, called the Graph Convolutional Network (GCN), has attracted much attention. It provides an effective mechanism to take into account the global information via the convolutional operation on a global graph and has achieved good results in many tasks including text classification. In this thesis, we propose a method that can combine both advantages of BERT model, which is excellent at exploiting the local information from a text, and the GCN model, which provides the application-dependent global language information. However, the traditional GCN is a transductive learning model, which performs a convolutional operation on a graph composed of task entities (i.e. documents graph) and cannot be applied directly to a new document. In this thesis, we first propose a novel Vocabulary GCN model (VGCN), which transforms the document-level convolution of the traditional GCN model to word-level convolution using a word graph created from word co-occurrences. In this way, we change the training method of GCN, from the transductive learning mode to the inductive learning mode, that can be applied to new documents. Secondly, we propose an Interactive-VGCN-BERT model that combines our VGCN model with BERT. In this model, local information including dependencies between words in a sentence, can be captured by BERT, while the global information reflecting the relations between words in a language (e.g. related words) can be captured by VGCN. In addition, local information and global information can interact through different layers of BERT, allowing them to influence mutually and to build together a final representation for classification. In so doing, the global language information can help distinguish ambiguous words or understand unclear expressions, thereby improving the performance of text classification tasks. To evaluate the effectiveness of our Interactive-VGCN-BERT model, we conduct experiments on several datasets of different types -- hate language detection, as well as movie review and grammaticality, and compare them with several state-of-the-art baseline models. Experimental results show that our Interactive-VGCN-BERT outperforms all other models such as Vanilla-VGCN-BERT, BERT, Bi-LSTM, MLP, GCN, and so on. In particular, we have found that VGCN can indeed help understand a text when it is integrated with BERT, confirming our intuition to combine the two mechanisms.