Capsule Networks are the next generation of image classifiers. Although they have several advantages over conventional Convolutional Neural Networks (CNNs), they remain computationally heavy. Since inference on Capsule Networks is timeconsuming, thier usage becomes limited to tasks in which latency is not essential. Approximation methods in Deep Learning help networks lose redundant parameters to increase speed and lower energy consumption. In the first part of this work, we go through an algorithm called zero-skipping. More than 50% of trained CNNs consist of zeros or values small enough to be considered zero. Since multiplication by zero is a trivial operation, the zero-skipping algorithm can play a massive role in speed increase throughout the network. We investigate the eligibility of Capsule Networks for this algorithm on two different datasets. Our results suggest that Capsule Networks contain enough zeros in their Primary Capsules to benefit from this algorithm. In the second part of this thesis, we investigate pruning as one of the most popular Neural Network approximation methods. Pruning is the act of finding and removing neurons which have low or no impact on the output. We run experiments on four different datasets. Pruning Capsule Networks results in the loss of redundant Primary Capsules. The results show a significant increase in speed with a minimal drop in accuracy. We also, discuss how dataset complexity affects the pruning strategy. / Graduate

Capsule Network (CapsNet) was introduced in 2017 as the new generation of the image classifiers to perform supervised classification of images. It incorporates a new structure of neurons which is called a capsule. A capsule is basically a vector of neurons and serves as the basic computation unit in CapsNet. CapsNet has obtained state-of-the-art testing accuracy on the task of classifying the MNIST digit recognition dataset. Despite its fundamental advantages over CNNs, it has its own shortcomings as well. CapsNet provides a relatively high accuracy in classifying images with affine transforms applied to them and also classifying images containing overlapping categories, compared to CNNs. Unlike CNNs, CapsNet creates the representation based on the part to whole relationship of the features of different levels. As a result, it comes with a more robust representation of the input image. CapsNet could only get reasonable inference accuracy on small-scale datasets. Also, it only supports a limited number of categories in the classification task. Finally, CapsNet is a relatively slow network, which is mostly due to the iterative Dynamic Routing (DR) algorithm used in it. There have been several works trying to address the shortcomings of CapsNet since it was introduced. In this work, we focus on optimizing CapsNet in several aspects: the network speed i.e. training and testing times, the number of parameters in the network, the network accuracy and its generalization ability. We propose several optimizations in order to compensate for the drawbacks of CapsNet. First, we introduce the Quick-CapsNet (QCN) network with our primary focus on the network speed. QCN makes changes to the feature extractor of CapsNet and produces fewer capsules compared to the baseline network (Base-CaspsNet). It performs inference 5x faster on small-scale datasets i.e. MNIST, F-MNIST, SVHN and CIFAR-10. QCN however loses testing accuracy marginally compared to the baseline e.g. 1% for F-MNIST dataset. Our second contribution is designing a capsule-specific layer for the feature extractor of CapsNet referred to as the Convolutional Fully-Connected (CFC) layer. We employ the CFC layer into CapsNet and call this new architecture CFC-CapsNet. CFC layer is added on top of the current feature extractor to translate the feature map into capsules. This layer has two parameters: kernel size and the output dimension. We performed some experiments to explore the effect of these two parameters on the network performance. Using the CFC layer results in reducing the number of parameters, faster training and testing, and higher test accuracy. On the CIFAR-10 dataset, CFC-CapsNet gets 1.46% higher accuracy (with baseline of 71.69%) and 49% fewer number of parameters. CFC-CapsNet is 4x and 4.5x faster than Base-CapsNet on CIFAR-10 for training and testing respectively. Our third contribution includes the introduction of LE-CapsNet as a light, enhanced and resource-aware variant of CapsNet. This network contains a Primary Capsule Generator (PCG) module as well as a robust decoder. Using 3.8M weights, LE-CapsNet obtains 77.21% accuracy for the CIFAR-10 dataset while performing inference 4x faster than CapsNet. In addition, our proposed network is more robust at detecting images with affine transformations compared to CapsNet. We achieve 94.37% accuracy on the AffNIST dataset (compared to CapsNet's 90.52%). Finally, we propose a deep variant of CapsNet consisting of several capsule layers referred to as Deep Light CapsNet (DL-CasNet). In this work, we design the Capsule Summarization (CapsSum) layer to reduce the complexity of the proposed deep network by reducing the number of parameters. DL-CapsNet, while being highly accurate, employs a small number of parameters compared to the state-of-the-art CapsNet based networks. Moreover DL-CapsNet delivers faster training and inference. Using a 7-ensemble model on the CIFAR-10 dataset, we achieve a 91.29% accuracy. DL-CapsNet is among the few networks based on CapsNet that supports the CIFAR-100 dataset (68.36% test accuracy using the 7-ensemble model) and can process complex datasets with a high number of categories. / Graduate

Scoliosis is a common skeletal condition in which a curvature forms along the coronal plane of the spine. Although scoliosis has been long recognized, its pathophysiology and best mode of treatment are still debated. Currently, definitive diagnosis of scoliosis and its progression are performed through anterior-posterior (AP) radiographs by measuring the angle of coronal curvature, referred to as Cobb angle. Cobb angle measurements can be performed by Deep Learning algorithms and are currently being investigated as a possible diagnostic tool for clinicians. This thesis focuses on the role of Deep Learning in the diagnosis and treatment of Scoliosis and proposes a study design using the algorithms to continue to better understand and classify the disease.

Thesis provides an overview and discussion of current findings in the field of biometrics. In particular, it focuses on facial recognition subject. Special attention is payed to convolutional neural networks and capsule networks. Thesis then lists current approaches and state-of-the-art implementations. Based on these findings it provides insight into engineering a very own solution based of CapsNet architecture. Moreover, thesis discussed advantages and capabilitied of capsule neural networks for identification of a person by its face.

Capsule network (CapsNet) is a new and promising approach to computer vision. In the small amount of research published so far, it has shown to be good at generalizing complex objects and perform well even when the images are skewed or the objects are seen from unfamiliar viewpoints. This thesis further tests this ability of CapsNetby comparing it to convolutional networks (ConvNets) on the task to understand images of clothing in different rotational views. Even though the ConvNets have a higher classification accuracy than CapsNets, the results indicate that CapsNets are better at understanding the clothes when viewed in different rotational views. / Capsule network (CapsNet) är en ny typ av neuralt nätverk för datorseende, som framförallt presterar bra även då bilderna är förvrängda eller sedda från obekanta vinklar. Den här uppsatsen testar CapsNets förmåga att förstå klädesobjekt sedda ur olika synviklar genom att göra en jämförelse med ConvNets. Resultaten visar att, trots att ConvNets har en högre exakthet i sin klassificering, är CapsNets bättre på att förstå kläderna sedda från olika synvinklar.