Physical Sciences and Mathematics Commons^™

Metric Learning Via Linear Embeddings For Human Motion Recognition, Byoungdoo Kong

Masters Theses

We consider the application of Few-Shot Learning (FSL) and dimensionality reduction to the problem of human motion recognition (HMR). The structure of human motion has unique characteristics such as its dynamic and high-dimensional nature. Recent research on human motion recognition uses deep neural networks with multiple layers. Most importantly, large datasets will need to be collected to use such networks to analyze human motion. This process is both time-consuming and expensive since a large motion capture database must be collected and labeled. Despite significant progress having been made in human motion recognition, state-of-the-art algorithms still misclassify actions because of characteristics …

Ppmexplorer: Using Information Retrieval, Computer Vision And Transfer Learning Methods To Index And Explore Images Of Pompeii, Cindy Roullet

Graduate Theses and Dissertations

In this dissertation, we present and analyze the technology used in the making of PPMExplorer: Search, Find, and Explore Pompeii. PPMExplorer is a software tool made with data extracted from the Pompei: Pitture e Mosaic (PPM) volumes. PPM is a valuable set of volumes containing 20,000 historical annotated images of the archaeological site of Pompeii, Italy accompanied by extensive captions. We transformed the volumes from paper, to digital, to searchable. PPMExplorer enables archaeologist researchers to conduct and check hypotheses on historical findings. We present a theory that such a concept is possible by leveraging computer generated correlations between artifacts using …

Dataset And Evaluation Of Self-Supervised Learning For Panoramic Depth Estimation, Ryan Nett

Master's Theses

Depth detection is a very common computer vision problem. It shows up primarily in robotics, automation, or 3D visualization domains, as it is essential for converting images to point clouds. One of the poster child applications is self driving cars. Currently, the best methods for depth detection are either very expensive, like LIDAR, or require precise calibration, like stereo cameras. These costs have given rise to attempts to detect depth from a monocular camera (a single camera). While this is possible, it is harder than LIDAR or stereo methods since depth can't be measured from monocular images, it has to …

Attentional Parsing Networks, Marcus Karr

Master's Theses

Convolutional neural networks (CNNs) have dominated the computer vision field since the early 2010s, when deep learning largely replaced previous approaches like hand-crafted feature engineering and hierarchical image parsing. Meanwhile transformer architectures have attained preeminence in natural language processing, and have even begun to supplant CNNs as the state of the art for some computer vision tasks.

This study proposes a novel transformer-based architecture, the attentional parsing network, that reconciles the deep learning and hierarchical image parsing approaches to computer vision. We recast unsupervised image representation as a sequence-to-sequence translation problem where image patches are mapped to successive layers …

The Wall: A Mobile App To Identify And Store Social Events From A Digital Image Using Computer Vision, Akhill Chandran, Ana Julia Ortiz, Eliezer Maia Barbosa, Maura Carola Tangara, Raquel Martini

ICT

Social events, promoted in print media using posters, flyers and banners often fail to attract an audience because we frequently forget the details of the event when we pass-by the promotion on the street. Smaller venues or artists often rely on low-cost, street-level marketing campaigns in areas of high foot traffic areas to develop interest in an event. These venues or artist are often without a budget for online marketing or have a target demographic outside the typical Social Media consumer which makes attracting an audience difficult.

This project aimed to solve the problem of storing and reminding the user …

Moving-Camera Video Content Analysis Via Action Recognition And Homography Transformation, Yang Mi

Theses and Dissertations

Moving-camera video content analysis aims at interpreting useful information in videos taken by moving cameras, including wearable cameras and handy cameras. It is an essential problem in computer vision, and plays an important role in many real-life applications, including understanding social difficulties and enhancing public security. In this work, we study three sub-problems of moving-camera video content analysis, including two sub-problems for the analysis on wearable-camera videos which are a special type of moving camera videos: recognizing general actions and recognizing microactions in wearable-camera videos. And, the third sub-problem is estimating homographies along moving-camera videos.

Recognizing general actions in wearable-camera …

Attacking Computer Vision Models Using Occlusion Analysis To Create Physically Robust Adversarial Images, Jacobsen Loh

Master's Theses

Self-driving cars rely on their sense of sight to function effectively in chaotic and uncontrolled environments. Thanks to recent developments in computer vision, specifically convolutional neural networks, autonomous vehicles have developed the ability to see at or above human-level capabilities, which in turn has allowed for rapid advances in self-driving cars. Unfortunately, much like humans being confused by simple optical illusions, convolutional neural networks are susceptible to simple adversarial inputs. As there is no overlap between the optical illusions that fool humans and the adversarial examples that threaten convolutional neural networks, little is understood as to why these adversarial examples …

Detection Of Mild Cognitive Impairment Using Diffusion Compartment Imaging, Matthew Jones

Master's Projects

The result of applying the Neurite Orientation Density and Dispersion Index (NODDI) algorithm to improve the prediction accuracy for patients diagnosed with MCI is reported. Calculations were carried out using a collection of 68 patients (34 control and 34 with MCI) gathered from the Alzheimer’s Disease Neuroimaging Initiative database (ADNI). Patient data includes the use of high-resolution Magnetic Resonance Images as with as Diffusion Tensor Imaging. A Linear Regression accuracy of 83% was observed using the added NODDI summary statistic: Orientation Dispersion Index (ODI). A statistically significant difference in groups was found between control patients and patients with MCI with …

Towards Multi-Modal Data Classification, Henry Ng

UNLV Theses, Dissertations, Professional Papers, and Capstones

A feature fusion multi-modal neural network (MMN) is a network that combines different modalities at the feature level to perform a specific task. In this paper, we study the problem of training the fusion procedure for MMN. A recent study has found that training a multi-modal network that incorporates late fusion produces a network that has not learned the proper parameters for feature extraction. These late fusion models perform very well during training but fall short to its single modality counterpart when testing. We hypothesize that jointly trained MMN have weight space that is too large for effective training. To …

Estimating Free-Flow Speed With Lidar And Overhead Imagery, Armin Hadzic

Theses and Dissertations--Computer Science

Understanding free-flow speed is fundamental to transportation engineering in order to improve traffic flow, control, and planning. The free-flow speed of a road segment is the average speed of automobiles unaffected by traffic congestion or delay. Collecting speed data across a state is both expensive and time consuming. Some approaches have been presented to estimate speed using geometric road features for certain types of roads in limited environments. However, estimating speed at state scale for varying landscapes, environments, and road qualities has been relegated to manual engineering and expensive sensor networks. This thesis proposes an automated approach for estimating free-flow …

Glacier Segmentation In Satellite Images For Hindu Kush Himalaya Region, Bibek Aryal

Open Access Theses & Dissertations

Climate change poses a risk to individuals whose livelihoods depend on the health of glacier ecosystems. Monitoring glaciers in the Himalayan Hindu Kush (HKH) region is of high importance especially when we consider the impact of recent climate change on them. Our work aims to provide an automated method to outline glaciers using machine learning techniques and publicly available remote sensing imagery.In this work, we present ways to delineate glaciers from Landsat-7 imagery using various machine learning and computer vision techniques. The multi-step methodology that we present in this work is generalizable across different types of satellite and overhead imagery, …

Attention Mechanism In Deep Neural Networks For Computer Vision Tasks, Haohan Li

Doctoral Dissertations

“Attention mechanism, which is one of the most important algorithms in the deep Learning community, was initially designed in the natural language processing for enhancing the feature representation of key sentence fragments over the context. In recent years, the attention mechanism has been widely adopted in solving computer vision tasks by guiding deep neural networks (DNNs) to focus on specific image features for better understanding the semantic information of the image. However, the attention mechanism is not only capable of helping DNNs understand semantics, but also useful for the feature fusion, visual cue discovering, and temporal information selection, which are …

Representation Learning With Adversarial Latent Autoencoders, Stanislav Pidhorskyi M.S.

Graduate Theses, Dissertations, and Problem Reports

A large number of deep learning methods applied to computer vision problems require encoder-decoder maps. These methods include, but are not limited to, self-representation learning, generalization, few-shot learning, and novelty detection. Encoder-decoder maps are also useful for photo manipulation, photo editing, superresolution, etc. Encoder-decoder maps are typically learned using autoencoder networks.
Traditionally, autoencoder reciprocity is achieved in the image-space using pixel-wise
similarity loss, which has a widely known flaw of producing non-realistic reconstructions. This flaw is typical for the Variational Autoencoder (VAE) family and is not only limited to pixel-wise similarity losses, but is common to all methods relying upon …