7903 research outputs found
Sort by
General Purpose Tuning Data Visualization
Get PDFThis project centers on the visualization of High-Performance Computing (HPC) data obtained from the GPTune website. GPTune serves as a valuable resource for HPC experiments, providing a wealth of performance data from tuning studies and optimization tasks. Our objective is to develop an advanced data visualization framework tailored to GPTune’s datasets. Utilizing state-of-the-art visualization techniques, we aim to create an interactive platform that allows users to explore, analyze, and derive insights from the diverse tuning experiments conducted on HPC systems. The visualization tool will facilitate the identification of optimal configurations, performance trends, and patterns within GPTune data, empowering researchers and practitioners to make informed decisions and optimize their HPC applications effectively. This project aims to enhance the accessibility and interpretability of GPTune data, contributing to the broader understanding of HPC performance optimization
Virtual Museum Tours of the de Saisset Museum
Get PDFOur project transforms the California Stories exhibit at the de Saisset Museum into an immersive virtual tour, integrating Matterport technology and Human-Computer-Interaction (HCI) principles. Matterport’s high resolution 360 degree camera allows us to capture a physical space in high definition into a 3D virtual space while Matterport’s software helps to facilitate the creation and editing for our physical space in a virtual environment. HCI principles focus on how systems should work when direct interaction with users is a key component. We apply the principles of HCI to create a virtual museum tour that offers visitors a culturally enriching, interactive experience, elevating the standard of virtual museum tours beyond simple walkthroughs to engage a wider audience
Waypoint Profiler
Get PDFOcean health monitoring is crucial for maintaining the health of the ocean ecosystem. Currently, divers are deployed to collect data manually, which is both time and resource-consuming. Additionally, this process poses significant dangers to the divers. Therefore, a more efficient method for collecting oceanic data is needed. This thesis describes the design of a novel autonomous marine vehicle, the waypoint profiler. Launched from shore with scientific sensors, it autonomously navigates to ocean locations of interest and dives to measure key ocean health markers. The system integrates subsystems for scientific sensing, health monitoring, structural integrity, communications, and navigation/control, tailored to meet stakeholder needs such as the Monterey Bay Aquarium Research Institute (MBARI), the US Army Corps of Engineers, and Occidental College. The Scientific sensing subsystem measures water temperature and captures water samples. The Health subsystem tracks battery levels and detects leaks. The Structural subsystem protects components and supports operation in various conditions. The Navigation and Control subsystem uses GPS and thrusters for precise movement. Extensive testing and validation were conducted to ensure the system\u27s performance and reliability. The results show a partial success of our vehicle\u27s ability to navigate to GPS waypoints and dive vertically to profile water columns. In the future, improvements can be made to the design of an internal charging system, eliminating the need to disassemble the vehicle to remove the batteries for charging. Another area for improvement is the cluster control capabilities, allowing one or more vehicles to be deployed and work collaboratively to complete tasks more efficiently
Residual Transformer Unet for Medical Image Segmentation
Get PDFWith the continuous development of deep learning theory in the field of medical images, information technology-assisted treatment methods represented by medical image segmentation technology can help doctors to quickly determine the shape and location of the lesions and improve the diagnosis efficiency of brain tumors. Based on deep learning technology, this thesis carries out related research work on MRI image segmentation. The main contents are as follows:
To begin with, acquire and prepare the brain tumor (MRI) image segmentation dataset from the official MICCAI Society website. This involves normalizing the images, cropping, and slicing, as well as scaling the data to ensure that the dataset meets the input specifications of the deep learning model. This paper presents a medical image segmentation method based on an improved Swin U-net. Initially, an atrous spatial pyramid pooling module is introduced at the end of the encoder to capture multi-scale features, allowing the model to effectively understand image at different scales and fully extract contextual information. Subsequently, in the encoder, the original blocks are replaced with residual Swin Transformer Blocks, and on the decoder side, replaced with deep residual convolution blocks. This replacement preserves the original information and alleviates the gradient vanishing issues. Lastly, an attention gate mechanism is introduced in the skip connections, enabling the model to focus more on important features within the feature map and suppress irrelevant information, thereby improving the model\u27s segmentation accuracy.
The experimental results show that the improved segmentation model reached a validation Intersection over Union (IoU) of 89.47%, an increase of 4.36% over the Swin U-net model, demonstrating that it can effectively enhance the accuracy of image segmentation and optimize the results of the original model
Generative Video Compression: Achieving High Perceptual Quality at Low Bitrates
Get PDFVideo data now fills people’s daily routines, both in their professional endeavors and leisure pursuits. These activities impose significant pressure on the Internet bandwidth. It is essential to develop efficient video coding techniques that compress video data at a low bit rate, while decoding visually appealing frames and saving transmission bandwidth.
In traditional video codecs, like Advanced Video Coding (AVC), High Efficiency Video Coding (HEVC), and Versatile Video Coding (VVC), techniques rooted in signal processing and information theory have long been dominant. Recent years have witnessed a surge in deep learning-based approaches for image and video compression. Among these methods, Generative Adversarial Network (GAN) has shown remarkable efficiency in compressing images at low bit rates while preserving high perceptual quality. This characteristic renders GAN an excellent choice for applications requiring low bit rates, such as video conferencing. The decoded images produced by GAN-based methods typically exhibit finer details, compared to non-GAN-based methods. Furthermore, through adversarial training, GAN-based approaches yield decoded frames that better align with the characteristics of the Human Visual System (HVS) compared to methods focused solely on minimizing pixel-wise differences between raw and decompressed frames. Currently, GAN-based techniques are being successfully deployed in video compression to enhance perceptual quality at low bit rates. This research devised three GAN-based frameworks for video compression:
The residue frame contains redundant energy of videos and truly little research directly applied GAN to residue-frame coding. To significantly reduce the signal energies of the videos, our first work developed a novel approach for inter-frame residue-coding-based video compression by leveraging GAN. It utilizes GAN to compress the I-frame and the residue of the P-frame. Empirical evaluations conducted on standard test sequences underscore the remarkable efficacy of our algorithm. Our methods reduced the average bpp by 30.58%, 47.63%, and 34.98% while increasing the average PSNR by 1.29 dB, 1.92 dB, and 1.44 dB when compared to GAN-Intra [6], CAE [84], and End-to-End [10] methods. Crucially, the adversarial training we adopted ensured that the decoded target frames are not only well-preserved in high perceptual quality but also consistent with HVS.
Most existing video coding work adopts CNN or RNN in the framework. Though effective in catching local features, it doesn’t model long-distance dependencies and extract non-local features thoroughly. To explore global correlations among sequences and inspired by the potential of GAN to compress frames at extremely low bit rates, our second work introduced a novel generative video compression (GVC) model with a transformer-based discriminator (TD) for P-frame, operating on a residue-coding-based paradigm. This model captures non-local correlations within video frames, thereby enhancing adversarial training. Moreover, our GVC model constraints not only bitrate-distortion base loss but also incorporate a discriminator-dependent feature loss and a perceptual loss. The experiments on test sequences demonstrate that our GVC provides superior performance in terms of perceptual quality, as evidenced by significantly lower FID and KID scores at lower bit rates. Notably, GVC demonstrated exceptional performance at remarkably low bitrates (0.036 bpp to 0.067 bpp), surpassing existing methods such as PLVC [105], RLVC [104], and x265. Comparative analysis reveals that models like x265 (LDP very fast), x265 (LDP default), and RLVC (MS-SSIM) need higher bit rates (1.25× to 2.43×) to achieve comparable results, but they still introduce noticeable blurriness and noise artifacts in decoded frames.
Though as simple and efficient as the residue coding method, the videos still contain uncompressed redundant information. To further explore the correlations among frames, reduce bit rates, and learn richer local and global features, our third work developed a contextual generative video compression method with transformers (CGVC-T) which adopts contextual coding to improve coding efficiency and applies GAN for perceptual quality enhancement. The integrated hybrid transformer-convolution structure in auto-encoders of CGVC-T enables the learning of both global and local features within video frames to eliminate temporal and spatial redundancy. Experimental results on HEVC, UVG, and MCL-JCV datasets show that our CGVC-T achieves superior perceptual quality in terms of lower FID, KID, and LPIPS scores, compared to state-of-the-art learned video codecs, industrial video codecs x264 and x265, as well as official reference software JM, HM, and VTM. Additionally, the probability distribution models we developed resulted in lower bitrates required for transmitting the compressed video. On average, with the anchor VTM, CGVC-T achieves 11.7%, 31.8%, and 32.5% BD-rate savings in terms of FID, KID, and LPIPS, respectively. CGVC-T also outperforms all compared learned video codecs in terms of perceptual DISTS scores with a 20.6% BD-rate savings which is more than other compared methods
Aero Avengers: AIAA Design, Build, and Fly
Get PDFThis thesis explores the experience of a Santa Clara University senior design team that created an unmanned aerial vehicle (UAV) intended to compete in the ’Design, Build, Fly!’ competition hosted by the American Institute of Aeronautics and Astronautics (AIAA). The six students worked together to secure funding for the project, formulate a preliminary design, and then manufacture the aircraft. Every step of this process was iterative, allowing the team to learn from their past attempts as the project’s timeline progressed. Upon obtaining competition rules and mission requirements, the team followed a design flow that brought the plane from a sketched design to a fully operable UAV. Determining the desired flight speed and taking into account the competition runway length, the team settled on a SD7062 NACA airfoil. This, along with a 1ft chord length and 5ft wingspan, satisfied the high lift requirement obtained from known competition parameters. Choosing balsa as the primary building materials was fairly straightforward when taking into account the limited budget and plethora of hobby related information available online and in literature. Using a semi-monocoque design, a computer-aided design (CAD) model was created and allowed the team to perform finite-element analysis (FEA) on the structure. FEA simulations validated the team’s design and ensured that the calculations done to determine various sizing requirements of structural members were valid when applying the expected loads. In terms of subsystems, the landing gear was retained from last year’s team which allowed the team to reduce costs. The system’s thrust comes from a 15x6E propeller connected to a 2000W 520Kv brushless DC motor powered by a 22.2V 4,500mAh LiPo battery. Servo motors powered the control surfaces, and these were connected to a different power supply, namely a 5S 6.0V 2,000mAh NiMh battery, as requested by competition requirements. Upon the final manufacturing of the craft, it was able to complete eight successful test flights, however, before the eighth flight, the wing’s leading edge cracked which prevented further flight testing. Due to time constraints, a solution to this problem was not implemented on this plane. Proposals to this solution are going to be shared with next year’s prospective team
Autonomous Hydroponic System
Get PDFHydroponics is a potentially useful farming technique for growing food in climates where it is difficult to naturally farm. However, hydroponic systems in harsh climates are usually very large and unable to be easily moved around. The design of a portable, enclosed hydroponic garden using the soilless Nutrient Film Technique (NFT) with a vertical configuration that can grow spinach in Antarctica will be discussed. It will have a reversible flow to evenly distribute nutrients to stacked rows of plants. A tradeoff analysis shows that 3D printing a custom T splitter is the best option for controlling the flow within the system. A SOLIDWORKS™ model was designed to show how the nutrient solution would flow through the system. ANSYS™ Granta EduPack was used to pick a material for insulation to help the garden retain heat and humidity. A tradeoff analysis shows that polyethylene terephthalate should be used to insulate the garden. When plants are growing inside the enclosure, conducting an energy analysis is a possible consideration for quantifying the efficiency of the garden
Introduction to Past and Future Presence: Approaches for Implementing XR Technology in Humanities and Art Education
Get PDFExtended reality (XR) technology (that is, virtual reality, augmented reality, mixed reality, and other 3D technologies) offers many opportunities and benefits for education in the humanities and arts. However, after attending many education- technology workshops and having read countless books and journal articles on the uses of XR in education over the past decade, it has been overwhelmingly apparent that most of the research and conference presentations tend to focus on the use of this technology in science, technology, engineering, and mathematics (STEM) education. While in many ways this is understandable, given the obvious applications of the technology to the fields and study of science, engineering, technology, and mathematics, the focus is still too limited and even to some extent exclusionary, failing to see and/ or acknowledge the potential benefits and creative opportunities in education outside of these areas of research and teaching. Additionally, though a multitude of digital humanities projects exist at many, if not all, institutes of higher education, the push for STEM education and careers in colleges and universities can be viewed as contributing to the lack of concentration on innovation in humanities education, as well as to declining enrollments in many non- STEM courses. Finally, for many scholars and teachers in fields outside of STEM, who are already incorporating this and similar types of technology into their teaching and research agendas, their work may not always be recognized or appreciated as scholarly effort in the same sense or worth as traditional research methods by those judging tenure cases, grant proposals, funding opportunities, and the like.
In this volume we aim to counteract some of the narrower views mentioned earlier through highlighting how XR technology can be used in (sometimes) less obvious but equally sophisticated and fruitful ways to create innovative, immersive, and interactive learning experiences for students in the arts and humanities.1 By presenting case studies from several arts and humanities fields at the undergraduate and graduate levels, our goals are to help: 1) inspire outside- the- box thinking; 2) move conversations from isolated silos into national and international discussions (and eventual collaborative projects) among researchers, scholars, teachers, and developers about how to best utilize this technology to complement and enhance current humanities teaching and learning practices in higher education and beyond; and 3) argue that performing research and creating teaching materials with this technology qualify as important scholarly endeavors and ought to be judged and valued as such by university administrators, faculty and other colleagues, grant committees, and more. This collection incorporates academic sources, project write- ups, and case studies that are intended to be used by instructors and administrators in secondary and post- secondary education to introduce or procure a better understanding of the benefits and drawbacks of XR technology in the humanities classroom. While academic in nature, this volume is also intended to be read as a practical document, and we encourage the reader to learn from the authors’ successes and mistakes which they encountered in the process of exploring this emerging medium
A Clinical Social Worker Reads Genesis 2:4b - 3:24: Enlisting Attachment Theory and Trauma Studies for an Alternative Interpretation of the Classic Genesis Story of the Fall
No full textThe story of Adam and Eve in Genesis 2:4b-3:24 has been extensively interpreted through a lens of crime and punishment, guilt, moral culpability, and the original sin of pride, becoming a fixed tale of “the fall” and expulsion from Eden. This dominant view, entrenched for centuries, particularly in Western Christianity, neglects alternative perspectives. Drawing from biblical studies, psycho-social research, theology, and spirituality, this interdisciplinary endeavor seeks to present an alternative reading that emphasizes original relationships and suffering over moralistic judgment. Inspired by arguments suggesting the traumatic origins of biblical texts and informed by clinical social work, the project employs attachment theory and trauma research to offer a fresh perspective on this familiar narrative.
The methodology integrates literary and narrative criticism with theories from various disciplines, challenging the dominant interpretation by embracing complementary hermeneutics of memory, hunger, and generosity toward the story and its characters. By adopting this approach, the project aims to excavate unconsidered possibilities within the text, inviting readers to engage with the story with renewed curiosity and compassion.
The chapter overview outlines the trajectory of the project, from defining the dominant narrative and tracing its evolution to moving toward a compelling reinterpretation of Genesis as a story of trauma and resilience. Each chapter contributes to the overarching goal of presenting Genesis 2-3 as text of trauma and resilience, resonating with contemporary concerns and enriching the spiritual journey of readers.
This project holds significance across multiple fields, contributing to ongoing conversations in biblical studies, trauma-sensitive approaches to Christian theology and praxis, and spirituality. It offers a novel perspective that can enhance understanding of self, others, and the divine, particularly for individuals who have experienced trauma. Liberating the narrative from constricting interpretations opens possibilities for readers to relate to scripture in fresh, transformative ways
Past and Future Presence: Approaches for Implementing XR Technology in Humanities and Art Education
No full textWhile uses and studies of XR technology within STEM-based education have been plentiful in recent years, there has been lesser or even, at times, a lack of coverage for this novel learning tool in the arts and humanities.Past and Future Presence aims to bridge some of that gap by presenting research-based theory and case studies of successful application and implementation of XR technology into postsecondary educational settings, ranging in topics from ancient to modern languages, classical and contemporary art, and reenvisioned historical scenes and events presented in ways never seen before. The studies also contemplate how this novel medium can enhance and supplement learning in classrooms and other formal or informal learning environments. The volume as a whole is intended to demonstrate to educators, scholars, and researchers in higher education the potential value of integrating XR technology into their classrooms and to provide a strong argument for college and university administrators to invest in training and development of new research and content for classrooms inside and outside of STEM. The authors of these chapters come from a diverse range of backgrounds at different stages of their careers, providing a broad crosssection of scholastic work within the humanities and arts. Each chapter offers a different angle or approach to incorporating XR technology into teaching or research within different subject areas. As the volume suggests, this technology also places additional emphasis on the humanity within the humanities, by focusing on increasing connection between users and different cultures, time periods, and perspectives.https://scholarcommons.scu.edu/faculty_books/1626/thumbnail.jp