Physical Sciences and Mathematics Commons^™

Automatic Music Transcription With Convolutional Neural Networks Using Intuitive Filter Shapes, Jonathan Sleep

Master's Theses

This thesis explores the challenge of automatic music transcription with a combination of digital signal processing and machine learning methods. Automatic music transcription is important for musicians who can't do it themselves or find it tedious. We start with an existing model, designed by Sigtia, Benetos and Dixon, and develop it in a number of original ways. We find that by using convolutional neural networks with filter shapes more tailored for spectrogram data, we see better and faster transcription results when evaluating the new model on a dataset of classical piano music. We also find that employing better practices shows …

Eulerian On Lagrangian Cloth Simulation, Kyle C. Piddington

Master's Theses

This thesis introduces a novel Eulerian-on-Lagrangian (EoL) approach for simulating cloth. This approach allows for the simulation of traditionally difficult cloth scenarios, such as draping and sliding cloth over sharp features like the edge of a table. A traditional Lagrangian approach models a cloth as a series of connected nodes. These nodes are free to move in 3d space, but have difficulty with sliding over hard edges. The cloth cannot always bend smoothly around these edges, as motion can only occur at existing nodes. An EoL approach adds additional flexibility to a Lagrangian approach by constructing special Eulerian on Lagrangian …

The Acquisition And Analysis Of Electroencephalogram Data For The Classification Of Benign Partial Epilepsy Of Childhood With Centrotemporal Spikes, Jessica A. Scarborough

Master's Theses

In this thesis, I will expand upon each step in the process of acquiring and analyzing electroencephalogram (EEG) for the classification of benign childhood epilepsy with centrotemporal spikes. Despite huge advancements in the field of health informatics—natural language processing, machine learning, predictive modeling—there are significant barriers to the access of clinical data. These barriers include information blocking, privacy policy concerns, and a lack of stakeholder support. We will see that these roadblocks are all responsible for stunting biomedical research in some way, including my own experiences in acquiring the data for the second chapter of this thesis.

This second chapter …

Probabilistic Roadmaps For Virtual Camera Pathing With Cinematographic Principles, Katherine Davis

Master's Theses

As technology use increases in the world and inundates everyday life, the visual aspect of technology or computer graphics becomes increasingly important. This thesis presents a system for the automatic generation of virtual camera paths for fly-throughs of a digital scene. The sample scene used in this work is an underwater setting featuring a shipwreck model with other virtual underwater elements such as rocks, bubbles and caustics. The digital shipwreck model was reconstructed from an actual World War II shipwreck, resting off the coast of Malta. Video and sonar scans from an autonomous underwater vehicle were used in a photogrammetry …

Gpumap: A Transparently Gpu-Accelerated Map Function, Ivan Pachev

Master's Theses

As GPGPU computing becomes more popular, it will be used to tackle a wider range of problems. However, due to the current state of GPGPU programming, programmers are typically required to be familiar with the architecture of the GPU in order to effectively program it. Fortunately, there are software packages that attempt to simplify GPGPU programming in higher-level languages such as Java and Python. However, these software packages do not attempt to abstract the GPU-acceleration process completely. Instead, they require programmers to be somewhat familiar with the traditional GPGPU programming model which involves some understanding of GPU threads and kernels. …

Real-Time Fall Detection And Response On Mobile Phones Using Machine Learning, Ilona Shparii

Master's Theses

Falls are common and often dangerous for groups with impaired mobility, like the elderly or people with lower limb amputations. Finding ways of minimizing the frequency or impact of a fall can improve quality of life dramatically. When someone does fall, real-time detection of the fall and a long-lie can trigger fast medical assistance. Such a system can also collect reliable data on the nature of real-world falls that can be used to better understand the circumstances, to aid in prevention efforts. This work has been to develop a real-time fall tracking system specifically for subjects with lower limb amputations. …

A Mobile App Illustrating Sensory Neural Coding Through An Efficient Coding Of Collected Images And Sounds, Xiaolu Zhao

Master's Theses

Sensory neuroscience in the early auditory and visual systems appears distinct not

only to outside observers, but to many trained neuroscientists as well. However, to a computational neuroscientist, both sensory systems represent an efficient neural coding of information. In fact, on a computational level it appears the brain is using the same processing strategy for both senses - the same algorithm with just a change in inputs. Insights like this can greatly simplify our understanding of the brain, but require a significant computational background to fully appreciate. How can such illuminating results of computational neuroscience be made more accessible to …