Physical Sciences and Mathematics Commons^™

Real–Time Semantic Segmentation For Railway Anomalies Analysis, Paul Stanik Iii

UNLV Theses, Dissertations, Professional Papers, and Capstones

In the past few years, computer vision has made huge jumps due to deep learning which leverages increased computational power and access to data. The computer vision community has also embraced transparency to accelerate research progress by sharing open datasets and open source code. Access to large scale datasets and benchmark challenges propelled and opened the field. The autonomous vehicle community is a prime example. While there has been significant growth in the automotive vision community, not much has been done in the rail domain. Traditional rail inspection methods require special trains that are run during down time, have sensitive …

Exploring The Latent Space Of Image Captioning Networks, Mikian J. Musser

UNLV Theses, Dissertations, Professional Papers, and Capstones

State-of-the-art image captioning models can successfully produce a diverse set of accurate captions. Previous research has focused on improving caption diversity while maintaining a high level of fidelity. We shift the focus from accuracy and diversity to controllability. We use a modified version of the traditional encoder-decoder network that allows the model to produce a meaningful and structured latent space. We then explore the latent space using several latent cartographic methods: lerp, slerp, analogy completion, attribute vector rotation, and interpolation graphs. Additionally, we discuss different categories of latent space and provide modifications for each of the cartographic methods. Finally, we …

Forecasting Pedestrian Trajectory Using Deep Learning, Arsal Syed

UNLV Theses, Dissertations, Professional Papers, and Capstones

In this dissertation we develop different methods for forecasting pedestrian trajectories. Complete understanding of pedestrian motion is essential for autonomous agents and social robots to make realistic and safe decisions. Current trajectory prediction methods rely on incorporating historic motion, scene features and social interaction to model pedestrian behaviors. Our focus is to accurately understand scene semantics to better forecast trajectories. In order to do so, we leverage semantic segmentation to encode static scene features such as walkable paths, entry/exits, static obstacles etc. We further evaluate the effectiveness of using semantic maps on different datasets and compare its performance with already …

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 …

Uas-Based Object Tracking Via Deep Learning, Marc Dinh

UNLV Theses, Dissertations, Professional Papers, and Capstones

Tracking is the task of identifying an object of interest and detect its position over time, and has numerous applications like surveillance, security and traffic control. In present times, unmanned aerial vehicles (UAV) have been more and more common which provides us with a new and less explored domain, with an ideal vantage point for surveillance and monitoring applications.. Aerial tracking is a particularly challenging task as it introduces new environmental variables such as rapid motion in 3D space. We propose a new deep learned tracker architecture catered to aerial tracking.

First, a study of six state-of-the-art deep learned trackers …

Fall Detection By Using Video, Robert J. Gripentog

UNLV Theses, Dissertations, Professional Papers, and Capstones

Cameras have become common in our society and as a result there is more video available today than ever before. While the video can be used for entertainment or possibly as storage it can also be used as a sensor capturing crucial information, The information captured can be put to all types of uses, but one particular use is to identify a fall. The importance of identifying a fall can be seen especially in the older population that is affected by falls every year. The falls experienced by the elderly are devastating as they can cause apprehension to normal life …