Unmanned Aerial Vehicles (UAVs) provides a platform for many automated tasks and with an ever increasing advances in computing, these tasks can be more complex. The use of UAVs is expanded in this thesis with the goal of Search and Rescue (SAR), where a UAV can assist fast responders to search for a lost person and relay possible search areas back to SAR teams. To identify a person from an aerial perspective, low-level Histogram of Oriented Gradients (HOG) feature descriptors are used over a segmented region, provided from thermal data, to increase classification speed. This thesis also introduces a dataset to support a Bird’s-Eye-View (BEV) perspective and tests the viability of low level HOG feature descriptors on this dataset. The low-level feature descriptors are known as Boosted Histogram of Oriented Gradients (BHOG) features, which discretizes gradients over varying sized cells and blocks that are trained with a Cascaded Gentle AdaBoost Classifier using our compiled BEV dataset. The classification is supported by multiple sensing modes with color and thermal videos to increase classification speed. The thermal video is segmented to indicate any Region of Interest (ROI) that are mapped to the color video where classification occurs. The ROI decreases classification time needed for the aerial platform by eliminating a per-frame sliding window. Testing reveals that with the use of only color data iv and a classifier trained for a profile of a person, there is an average recall of 78%, while the thermal detection results with an average recall of 76%. However, there is a speed up of 2 with a video of 240x320 resolution. The BEV testing reveals that higher resolutions are favored with a recall rate of 71% using BHOG features, and 92% using Haar-Features. In the lower resolution BEV testing, the recall rates are 42% and 55%, for BHOG and Haar-Features, respectively.

A logistics centre, or control centre, is a major component for many industries, especially those that participate in activities related to transport vehicles and logistics, for instance, the mining industry. Previous research has shown that the capabilities of humans to process a large amount of information are limited. As the flow of information continues to increase in the future, Situation Awareness (SA) and task performance can decrease. Moreover, currently, an operator is dependent and restricted to their workstation to perform any necessary task, which is also a major hindrance that is being faced. This thesis project used Augmented Reality (AR) to tackle these problems and look at how the user experience of logistics operators working in these control centres could look like and improve in the next few decades. An augmented reality multimodal prototype was developed using a variety of technologies during this thesis project. The prototype was evaluated through both qualitative and quantitative methods. Six subject matter experts (SMEs) were selected as participants and completed a small quantitative survey, after which they were interviewed for more in-depth feedback. Results show a positive approval rating for the prototype, both in terms of usability and learnability. Additionally, a number of further solutions and possible applications emerged from interviews with the participants. Finally, it became apparent that one of the research sub-questions, about the overflow of information, could not be fully addressed by this prototype. However, possible solutions to that problem emerged from the project. / Ett logistikcenter eller kontrollcenter är en viktig komponent för många industrier, särskilt de som deltar i aktiviteter relaterade till transportfordon och logistik, till exempel gruvindustrin. Tidigare forskning har visat att människors förmåga att bearbeta en stor mängd information är begränsad. Eftersom informationsflödet fortsätter att öka i framtiden kan Situation Awareness (SA) och uppgiftsprestanda minska. Dessutom är en operatör för närvarande beroende och begränsad till sin arbetsstation för att utföra alla nödvändiga uppgifter, vilket också är ett stort hinder som står inför. Detta avhandlingsprojekt använde Augmented Reality (AR) för att ta itu med dessa problem och titta på hur användarupplevelsen för logistikoperatörer som arbetar i dessa kontrollcentra kan se ut och förbättras under de närmaste decennierna. En förstärkt verklighet multimodal prototyp utvecklades med hjälp av en mängd olika tekniker under detta avhandlingsprojekt. Prototypen utvärderades genom både kvalitativa och kvantitativa metoder. Sex ämnesexperter (SMF) valdes ut som deltagare och genomförde en liten kvantitativ undersökning, varefter de intervjuades för mer ingående feedback. Resultaten visar ett positivt godkännandebetyg för prototypen, både när det gäller användbarhet och lärbarhet. Dessutom framkom ett antal ytterligare lösningar och möjliga tillämpningar från intervjuer med deltagarna. Slutligen blev det uppenbart att en av forskningsundersökningarna, om överflöd av information, inte helt kunde behandlas av denna prototyp. Men möjliga lösningar på det problemet framkom ur projektet.

No description available.

No description available.

No description available.

No description available.

No description available.

Deep Learning techniques are widely used across various medical imaging applications. However, they are often fine-tuned for a specific modality and are not generalizable when it comes to new modalities or datasets. One of the main reasons for this is large data variations for e.g., the dynamic range of intensity values is large across multi-modal images. The goal of the project is to develop a method to address multi-modal learning that aims at segmenting liver from Computed Tomography (CT) images and abdominal organs from Magnetic Resonance (MR) images using deep learning techniques. In this project, a self-supervised approach is adapted to attain domain adaptation across images while retaining important 3D information from medical images using a simple 3D-UNet with a few auxiliary tasks. The method comprises of two main steps: representation learning via self-supervised learning (pre-training) and fully supervised learning (fine-tuning). Pre-training is done using a 3D-UNet as a base model along with some auxiliary data augmentation tasks to learn representation through texture, geometry and appearances. The second step is fine-tuning the same network, without the auxiliary tasks, to perform the segmentation tasks on CT and MR images. The annotations of all organs are not available in both modalities. Thus the first step is used to learn general representation from both image modalities; while the second step helps to fine-tune the representations to the available annotations of each modality. Results obtained for each modality were submitted online, and one of the evaluations obtained was in the form of DICE score. The results acquired showed that the highest DICE score of 0.966 was obtained for CT liver prediction and highest DICE score of 0.7 for MRI abdominal segmentation. This project shows the potential to achieve desired results by combining both self and fully-supervised approaches.

Master of Science / There has been an astronomical increase in data at the network edge due to the rapid development of 5G infrastructure and the proliferation of the Internet of Things (IoT). In order to improve the network controller's decision-making capabilities and improve the user experience, it is of paramount importance to properly analyze this data. However, transporting such a large amount of data from edge devices to the network controller requires large bandwidth and increased latency, presenting a significant challenge to resource-constrained wireless networks. By using information processing techniques, one could effectively address this problem by sending only pertinent and critical information to the network controller. Nevertheless, finding critical information from high-dimensional observation is not an easy task, especially when large amounts of background information are present. Our thesis proposes to extract critical but low-dimensional information from high-dimensional observations using an information-theoretic deep learning framework. We focus on two distinct problems where critical information extraction is imperative. In the first problem, we study the problem of feature extraction from video frames collected in a dynamic environment and showcase its effectiveness using a video game simulation experiment. In the second problem, we investigate the detection of anomaly signals in the spectrum by extracting and analyzing useful features from spectrograms. Using extensive simulation experiments based on a practical data set, we conclude that our proposed approach is highly effective in detecting anomaly signals in a wide range of signal-to-noise ratios.

<p dir="ltr">The ability to predict and manipulate how a particulate material will flow in a process is challenging for industry and researchers alike. This dissertation presents the results of a model-directed, experimental approach using a concentric cylinder rheometer titled along an axis to enable continuous chute flow of granular media. Experiments were performed using draining flows for constant and oscillatory applied shear rates. Multiple flow and stress sensors were used to investigate the interaction of mass holdup, shear rate, specific torque, particle velocity, discharge mass flow rate, and wall pressure. Depending on the flow configuration, linear ranges were observed wherein the specific torque remained steady during draining. This finding enabled systematic testing of flow behavior as a function of dimensionless shear rates. Results suggest changes in the specific torque, wall slip, and outflow variance occur with the transition from the quasi-static to dense-inertial flow regimes. A pump-curve analogy was also identified for the relationship between the outlet mass flow rate and the specific power relationship for the constant shear rate experiments. Oscillatory shear rate experiments show a significant influence of the phase shift between the applied shear rate and the specific torque. Adding an asperity to the rotor revealed rate-dependent patterns in bulk flow and force chain dynamics. Overall, the study offers valuable insights into the effects of shear rate and boundary conditions on dense granular flows. The effects of particle characteristics (e.g., size and shape distributions, friction, cohesivity) and material properties (e.g., density, modulus) remain topics for future work. </p>