Computer Sciences Commons^™

Object Detection With Deep Learning To Accelerate Pose Estimation For Automated Aerial Refueling, Andrew T. Lee

Theses and Dissertations

Remotely piloted aircraft (RPAs) cannot currently refuel during flight because the latency between the pilot and the aircraft is too great to safely perform aerial refueling maneuvers. However, an AAR system removes this limitation by allowing the tanker to directly control the RP A. The tanker quickly finding the relative position and orientation (pose) of the approaching aircraft is the first step to create an AAR system. Previous work at AFIT demonstrates that stereo camera systems provide robust pose estimation capability. This thesis first extends that work by examining the effects of the cameras' resolution on the quality of pose …

Maximizing Accuracy Through Stereo Vision Camera Positioning For Automated Aerial Refueling, Kirill A. Sarantsev

Theses and Dissertations

Aerial refueling is a key component of the U.S. Air Force strategic arsenal. When two aircraft interact in an aerial refueling operation, the accuracy of relative navigation estimates are critical for the safety, accuracy and success of the mission. Automated Aerial Refueling (AAR) looks to improve the refueling process by creating a more effective system and allowing for Unmanned Aerial Vehicle(s) (UAV) support. This paper considers a cooperative aerial refueling scenario where stereo cameras are used on the tanker to direct a \boom" (a large, long structure through which the fuel will ow) into a port on the receiver aircraft. …

Domain Adaptation For Vehicle Detection In Traffic Surveillance Images From Daytime To Nighttime, Jinlong Ji, Zhigang Xu, Hongkai Yu, Lan Fu, Xuesong Zhou

Computer Science Faculty Publications

Vehicle detection in traffic surveillance images is an important approach to obtain vehicle data and rich traffic flow parameters. Recently, deep learning based methods have been widely used in vehicle detection with high accuracy and efficiency. However, deep learning based methods require a large number of manually labeled ground truths (bounding box of each vehicle in each image) to train the Convolutional Neural Networks (CNN). In the modern urban surveillance cameras, there are already many manually labeled ground truths in daytime images for training CNN, while there are little or much less manually labeled ground truths in nighttime images. In …

Use Of Lidar In Automated Aerial Refueling To Improve Stereo Vision Systems, Michael R. Crowl

Theses and Dissertations

The United States Air Force (USAF) executes five Core Missions, four of which depend on increased aircraft range. To better achieve global strike and reconnaissance, unmanned aerial vehicles (UAVs) require aerial refueling for extended missions. However, current aerial refueling capabilities are limited to manned aircraft due to technical difficulties to refuel UAVs mid-flight. The latency between a UAV operator and the UAV is too large to adequately respond for such an operation. To overcome this limitation, the USAF wants to create a capability to guide the refueling boom into the refueling receptacle. This research explores the use of light detection …

Monocular Depth Image Mark-Less Pose Estimation Based On Feature Regression, Chen Ying, Shen Li

Journal of System Simulation

Abstract: Monocular camera mark-less pose estimation system suffers low accuracy, robustness and efficiency due to variety of action, self-occlusion of human body. A method of feature exaction from point clouds was proposed, in which a single-to-multiple (S2M) feature regressor and a joint position regressor were designed to quickly and accurately predict the 3D positions of body joints from a single depth image without any temporal information. Experiment result shows that the estimation accuracy is superior to that of state-of-the-arts and multi-camera based methods.

Novel View Synthesis In Time And Space, Simon Niklaus

Dissertations and Theses

Novel view synthesis is a classic problem in computer vision. It refers to the generation of previously unseen views of a scene from a set of sparse input images taken from different viewpoints. One example of novel view synthesis is the interpolation of views in between the two images of a stereo camera. Another classic problem in computer vision is video frame interpolation, which is important for video processing. It refers to the generation of video frames in between existing ones and is commonly used to increase the frame rate of a video or to match the frame rate to …

A Systematic Literature Survey Of Unmanned Aerial Vehicle Based Structural Health Monitoring, Sreehari Sreenath

Theses, Dissertations and Capstones

Unmanned Aerial Vehicles (UAVs) are being employed in a multitude of civil applications owing to their ease of use, low maintenance, affordability, high-mobility, and ability to hover. UAVs are being utilized for real-time monitoring of road traffic, providing wireless coverage, remote sensing, search and rescue operations, delivery of goods, security and surveillance, precision agriculture, and civil infrastructure inspection. They are the next big revolution in technology and civil infrastructure, and it is expected to dominate more than $45 billion market value. The thesis surveys the UAV assisted Structural Health Monitoring or SHM literature over the last decade and categorize UAVs …

Automated Recognition Of Facial Affect Using Deep Neural Networks, Behzad Hasani

Electronic Theses and Dissertations

Automated Facial Expression Recognition (FER) has been a topic of study in the field of computer vision and machine learning for decades. In spite of efforts made to improve the accuracy of FER systems, existing methods still are not generalizable and accurate enough for use in real-world applications. Many of the traditional methods use hand-crafted (a.k.a. engineered) features for representation of facial images. However, these methods often require rigorous hyper-parameter tuning to achieve favorable results.

Recently, Deep Neural Networks (DNNs) have shown to outperform traditional methods in visual object recognition. DNNs require huge data as well as powerful computing units …

Design Of A Novel Wearable Ultrasound Vest For Autonomous Monitoring Of The Heart Using Machine Learning, Garrett G. Goodman

Browse all Theses and Dissertations

As the population of older individuals increases worldwide, the number of people with cardiovascular issues and diseases is also increasing. The rate at which individuals in the United States of America and worldwide that succumb to Cardiovascular Disease (CVD) is rising as well. Approximately 2,303 Americans die to some form of CVD per day according to the American Heart Association. Furthermore, the Center for Disease Control and Prevention states that 647,000 Americans die yearly due to some form of CVD, which equates to one person every 37 seconds. Finally, the World Health Organization reports that the number one cause of …

An Automated Method For Detecting Water Levels Using Computer Vision And Artificial Intelligence, Priyanjani Chowdary Chandra

Graduate Research Theses & Dissertations

Flooding is one of the most dangerous weather events today. Between 2015-2019, on average, it has caused more than 130 deaths every year in the USA alone. World Health Organization has reported that, between 1998-2017, floods have affected more than 2 billion people worldwide. The devastating nature of flood necessitates the continuous monitoring of water level in the rivers and streams in flood-prone areas to detect the incoming flood. In this thesis, we have designed and implemented a computer vision and AI-based system that continuously detect the water level in the creek. Our solution employs an effective template matching algorithm …

Computer Vision Gesture Recognition For Rock Paper Scissors, Nicholas Hunter

Senior Independent Study Theses

This project implements a human versus computer game of rock-paper-scissors using machine learning and computer vision. Player’s hand gestures are detected using single images with the YOLOv3 object detection system. This provides a generalized detection method which can recognize player moves without the need for a special background or lighting setup. Additionally, past moves are examined in context to predict the most probable next move of the system’s opponent. In this way, the system achieves higher win rates against human opponents than by using a purely random strategy.

Deep Temporal Motion Descriptor (Dtmd) For Human Action Recognition, Nudrat Nida, Muhammad Haroon Yousaf, Aun Irtaza, Sergio A. Velastin

Turkish Journal of Electrical Engineering and Computer Sciences

Spatiotemporal features have significant importance in human action recognition, as they provide the actor's shape and motion characteristics specific to each action class. This paper presents a new deep spatiotemporal human action representation, the deep temporal motion descriptor (DTMD), which shares the attributes of holistic and deep learned features. To generate the DTMD descriptor, the actor?s silhouettes are gathered into single motion templates by applying motion history images. These motion templates capture the spatiotemporal movements of the actor and compactly represent the human actions using a single 2D template. Then deep convolutional neural networks are used to compute discriminative deep …

An Application Of Deep Learning Models To Automate Food Waste Classification, Alejandro Zachary Espinoza

Dissertations and Theses

Food wastage is a problem that affects all demographics and regions of the world. Each year, approximately one-third of food produced for human consumption is thrown away. In an effort to track and reduce food waste in the commercial sector, some companies utilize third party devices which collect data to analyze individual contributions to the global problem. These devices track the type of food wasted (such as vegetables, fruit, boneless chicken, pasta) along with the weight. Some devices also allow the user to leave the food in a kitchen container while it is weighed, so the container weight must also …

Learning Robot Manipulation Tasks Via Observation, Michail Theofanidis

Computer Science and Engineering Dissertations - Archive

The coexistence of humans and robots has been the aspiration of many scientific endeavors in the past century. Most anthropomorphic or industrial robots are highly articulated and complex machines, which are designed to carry out tasks that often involve the manipulation of physical objects. Traditionally, robots learn how to perform such tasks with the aid of a human programmer or operator. In this regard, the human acts as a teacher who provides a demonstration of a task. From the data of the demonstration, the robot must learn a state-action mapping that accomplishes the task. This state-action mapping is often addressed …

Detecting Phone-Related Pedestrian Behavior Using A Two-Branch Convolutional Neural Network, Humberto Saenz

Theses and Dissertations

With the wide use of smart phones, distraction has become a major safety concern to roadway users. The distracted phone-use behaviors among pedestrians, like Texting, Game Playing and Phone Calls, have caused increasing fatalities and serious injuries. With the increasing usage of driver monitor systems on intelligent vehicles, distracted driver behaviors can be efficiently detected and warned. However, the research of phone-related distracted behavior by pedestrians has not been systemically studied. It is desired to improve both the driving and pedestrian safety by automatically discovering the phone-related pedestrian distracted behaviors. In this thesis, we propose a new computer vision-based method …

Learning To Detect Pedestrians By Watching Videos, Andrew Y. Chen

Theses and Dissertations

The field of deep learning has experienced a resurgence in the recent years, particularly resulting with the advent of AlexNet. Supervised learning is currently the most common and practical machine learning method. The struggle with employing supervised learning to approach problems is that it requires training data. Sufficient training data is correlated with performance for deep learning models. The issue is that preparing the training data can be a tedious and labor intensive task, especially on a large scale. The purpose of this paper is to determine how efficient a machine can learn when trained on automatically annotated data. The …

Detecting Digitally Forged Faces In Online Videos, Neilesh Sambhu

USF Tampa Graduate Theses and Dissertations

We use Rossler’s FaceForensics dataset of 1004 online videos and their corresponding forged counterparts [1] to investigate the ability to distinguish digitally forged facial images from original images automatically with deep learning. The proposed convolutional neural network is much smaller than the current state-of-the-art solutions. Nevertheless, the network maintains a high level of accuracy (99.6%), all while using the entire FaceForensics dataset and not including any temporal information. We implement majority voting and show the impact on accuracy (99.67%), where only 1 video of 300 is misclassified. We examine why the model misclassified this one video. In terms of tuning …

Pose Based Human Activity Recognition, Wenbo Li

Legacy Theses & Dissertations (2009 - 2024)

Pose based human activity recognition is an important step towards video understanding. The last decade has witnessed the great progress in this field which is driven by multiple technical innovations, i.e., kinect, pose estimation techniques, deep learning, etc.

Poriferal Vision, Saketh Saxena

Master's Projects

Sponges provide nourishment as well as a habitat for various aquatic organisms. Anatomically, sponges are made up of soft tissue with a silica based exoskeleton which serves both as support and protection for the underlying tissue. The exoskeleton persists after the tissue decomposes, and microscopic parts of the exoskeleton break away to form spicules. Oceanographic studies have shown that the density of the sponge spicules is a good indicator of the sponge population in an area. This measure can be used to study sponge population dynamics over time. The spicule density is measured by imaging spicules from samples of water …

Using Computer Vision To Quantify Coral Reef Biodiversity, Niket Bhodia

Master's Projects

The preservation of the world’s oceans is crucial to human survival on this planet, yet we know too little to begin to understand anthropogenic impacts on marine life. This is especially true for coral reefs, which are the most diverse marine habitat per unit area (if not overall) as well as the most sensitive. To address this gap in knowledge, simple field devices called autonomous reef monitoring structures (ARMS) have been developed, which provide standardized samples of life from these complex ecosystems. ARMS have now become successful to the point that the amount of data collected through them has outstripped …

Over Speed Detection Using Artificial Intelligence, Samkit Patira

Master's Projects

Over speeding is one of the most common traffic violations. Around 41 million people are issued speeding tickets each year in USA i.e one every second. Existing approaches to detect over- speeding are not scalable and require manual efforts. In this project, by the use of computer vision and artificial intelligence, I have tried to detect over speeding and report the violation to the law enforcement officer. It was observed that when predictions are done using YoloV3, we get the best results.

Classification Of Humans Into Ayurvedic Prakruti Types Using Computer Vision, Gayatri Gadre

Master's Projects

Ayurveda, a 5000 years old Indian medical science, believes that the universe and hence humans are made up of five elements namely ether, fire, water, earth, and air. The three Doshas (Tridosha) Vata, Pitta, and Kapha originated from the combinations of these elements. Every person has a unique combination of Tridosha elements contributing to a person’s ‘Prakruti’. Prakruti governs the physiological and psychological tendencies in all living beings as well as the way they interact with the environment. This balance influences their physiological features like the texture and colour of skin, hair, eyes, length of fingers, the shape of the …

Structured Indoor Modeling, Chen Liu

McKelvey School of Engineering Graduate Student Theses & Dissertations

In this dissertation, we propose data-driven approaches to reconstruct 3D models for indoor scenes which are represented in a structured way (e.g., a wall is represented by a planar surface and two rooms are connected via the wall). The structured representation of models is more application ready than dense representations (e.g., a point cloud), but poses additional challenges for reconstruction since extracting structures requires high-level understanding about geometries. To address this challenging problem, we explore two common structural regularities of indoor scenes: 1) most indoor structures consist of planar surfaces (planarity), and 2) structural surfaces (e.g., walls and floor) can …

Triplet Loss Network For Unsupervised Domain Adaptation, Imad Eddine Ibrahim Bekkouch, Youssef Youssry, Rustam Gafarov, Adil Khan, Asad Masood Khattak

All Works

© 2019 by the authors. Domain adaptation is a sub-field of transfer learning that aims at bridging the dissimilarity gap between different domains by transferring and re-using the knowledge obtained in the source domain to the target domain. Many methods have been proposed to resolve this problem, using techniques such as generative adversarial networks (GAN), but the complexity of such methods makes it hard to use them in different problems, as fine-tuning such networks is usually a time-consuming task. In this paper, we propose a method for unsupervised domain adaptation that is both simple and effective. Our model (referred to …

Real Time Facial Expression Recognition And Eye Gaze Estimation System, Suzan Anwar

Theses and Dissertations

The following dissertation proposes a gaze estimation and an emotion recognition system, as this system proposed is a face analysis package which not only contains face detection, but also eye tracking, emotion recognition, eye detection, as well as eye estimation. The system is comprised of a facial emotion recognition that can recognize the seven emotions including happiness, anger, sadness, neutral, surprise, disgust, and fear. This part of the system has an implemented Active Shape Model (ASM) tracker, as it is the emotion recognition part of the system, which through the webcam input can track 116 facial landmarks points to obtain …

Learning To Map The Visual And Auditory World, Tawfiq Salem

Theses and Dissertations--Computer Science

The appearance of the world varies dramatically not only from place to place but also from hour to hour and month to month. Billions of images that capture this complex relationship are uploaded to social-media websites every day and often are associated with precise time and location metadata. This rich source of data can be beneficial to improve our understanding of the globe. In this work, we propose a general framework that uses these publicly available images for constructing dense maps of different ground-level attributes from overhead imagery. In particular, we use well-defined probabilistic models and a weakly-supervised, multi-task training …

Recognition Of Incomplete Objects Based On Synthesis Of Views Using A Geometric Based Local-Global Graphs, Michael Christopher Robbeloth

Browse all Theses and Dissertations

The recognition of single objects is an old research field with many techniques and robust results. The probabilistic recognition of incomplete objects, however, remains an active field with challenging issues associated to shadows, illumination and other visual characteristics. With object incompleteness, we mean missing parts of a known object and not low-resolution images of that object. The employment of various single machine-learning methodologies for accurate classification of the incomplete objects did not provide a robust answer to the challenging problem. In this dissertation, we present a suite of high-level, model-based computer vision techniques encompassing both geometric and machine learning approaches …

Computer Vision-Based Traffic Sign Detection And Extraction: A Hybrid Approach Using Gis And Machine Learning, Zihao Wu

College of Graduate Studies: Theses & Dissertations

Traffic sign detection and positioning have drawn considerable attention because of the recent development of autonomous driving and intelligent transportation systems. In order to detect and pinpoint traffic signs accurately, this research proposes two methods. In the first method, geo-tagged Google Street View images and road networks were utilized to locate traffic signs. In the second method, both traffic signs categories and locations were identified and extracted from the location-based GoPro video. TensorFlow is the machine learning framework used to implement these two methods. To that end, 363 stop signs were detected and mapped accurately using the first method (Google …

Doppler Radar-Based Non-Contact Health Monitoring For Obstructive Sleep Apnea Diagnosis: A Comprehensive Review, Vinh Phuc Tran, Adel Ali Al-Jumaily, Syed Mohammed Shamsul Islam

Research outputs 2014 to 2021

Today’s rapid growth of elderly populations and aging problems coupled with the prevalence of obstructive sleep apnea (OSA) and other health related issues have affected many aspects of society. This has led to high demands for a more robust healthcare monitoring, diagnosing and treatments facilities. In particular to Sleep Medicine, sleep has a key role to play in both physical and mental health. The quality and duration of sleep have a direct and significant impact on people’s learning, memory, metabolism, weight, safety, mood, cardio-vascular health, diseases, and immune system function. The gold-standard for OSA diagnosis is the overnight sleep monitoring …

Elimination Of Useless Images From Raw Camera-Trap Data, Ulaş Tekeli̇, Yalin Baştanlar

Turkish Journal of Electrical Engineering and Computer Sciences

Camera-traps are motion triggered cameras that are used to observe animals in nature. The number of images collected from camera-traps has increased significantly with the widening use of camera-traps thanks to advances in digital technology. A great workload is required for wild-life researchers to group and label these images. We propose a system to decrease the amount of time spent by the researchers by eliminating useless images from raw camera-trap data. These images are too bright, too dark, blurred, or they contain no animals. To eliminate bright, dark, and blurred images we employ techniques based on image histograms and fast …