Scholarship RepositoryThe Florida Tech Scholarship Repository system captures, stores, indexes, preserves, and distributes digital research material.https://repository.lib.fit.edu:4432017-05-24T23:26:51Z2017-05-24T23:26:51ZAnalysis, A Technique, and Incremental Learning of Wake-Up-Word Speech RecognitionHasanain, Ahmad Zuhair S.http://hdl.handle.net/11141/14492017-05-24T20:43:21Z2017-05-01T00:00:00ZAnalysis, A Technique, and Incremental Learning of Wake-Up-Word Speech Recognition
Hasanain, Ahmad Zuhair S.
Even thought the cutting-edge speaker-independent Automatic Speech Recognition (ASR) systems
demand big training data, they barely handle time-varying speaking rates, tolerate various
uttering alterations, or are robust to noise. In contrast, our Wake-Up-Word (WUW) technique is
tuned to these challenges in the light ASR systems with minimal number of initial training samples.
It is crucial that users of ASR systems be capable of rolling out new WUW calls swiftly and
modifying ASR vocabulary at any time, such as in the cases of foreign WUW addition and adaptation
to phonetic change. We had tested our proposed methodologies in the acoustic WUW-II
corpus [17], and they guaranteed roughly 89% (±0.5%) for both Out-Of-Vocabulary (OOV) and
In-Vocabulary (INV) word recognition rates. We recommend dual directional (Bidirectional) Dynamic
Time Warping (BDTW), a chronological contrast model, and a semi-supervised training
procedure. Not only can BDTW produce accurate time alignment of phonemic states, but it can
also be utilized for WUW isolation, whereby boundaries of similar sounding patterns are precisely
located for autonomous segmenting/retrieval of WUWs from continuous speech streams. A
suggested distance/similarity model extracts time warping from the contrast of phones comprising
WUWs themselves by exploiting acoustic evidence up front. Hence, minimal prior knowledge
about language is needed with the proposed solutions. Additionally, just one utterance is capable
of initiating speaker-independent ASR systems when incremental learning is enabled after each
test such that the cognitive matching utterance of a speaker is the most probable hypothesis given
that each session contains at least one similar utterance. This work also conveys an overview
and analysis of fundamentals, implementations, and most importantly, results of empirical tests
and future focus.
Thesis (M.S.) - Florida Institute of Technology, 2017
2017-05-01T00:00:00ZCFD Analysis of Resistance Characteristics of High-Speed Displacement Hull Forms fitted with Hull Vane®Avala, Venkata Karthikhttp://hdl.handle.net/11141/14482017-05-24T20:05:44Z2017-05-01T00:00:00ZCFD Analysis of Resistance Characteristics of High-Speed Displacement Hull Forms fitted with Hull Vane®
Avala, Venkata Karthik
Improving the performance and resistance of a high-speed monohull forms has
been of interest to Naval Architects for several decades. A considerable amount
of research has been carried out in this area by using stern flaps and other
appendages.
This research thesis investigates the resistance characteristics of high-speed
round bilge hull forms fitted with a Hull Vane® in the stern region of the vessel.
The Hull Vane® is a fixed foil located below the waterline at the aft of the stern
of the vessel. The Hull Vane® reduces the generation of waves at the aft and
improves the vessel's motions. This thesis investigates the drag reduction by
fitting a Hull Vane® in the stern region of a select few of Australian Maritime
Engineering Cooperative Research Center (AMECRC) systematic series hull
forms, originally developed as High-Speed Displacement Hull Forms (HSDHF)
series at MARIN, The Netherlands.
The aim of this research is to:
- Validate the total resistance of AMECRC series models #3, #4, #8, #11,
#13 for the Froude numbers 0.5, 0.6, 0.7 in CFD against the experimental
data available. - Attach a Hull Vane® of NACA 4412 profile at the aft region of the above
hull forms at the same Froude numbers and determine reduction in resistance
if any.
Hull Vane® used for this analysis has a profile similar to NACA 4412 and principal
dimensions such as span has been varied according to the breadth of the
models tested. The flap angle and chord length have been kept constant. The CFD research is carried out on STAR CCM+, a CFD package used to
simulate the drag characteristics. Models #8, #11 and #13 have shown a
significant reduction in the total resistance varying from 7% to 29% with an
average reduction of 17.3% when fitted with a Hull Vane® . Model #3 and
#4 did not show any significant improvement in the resistance. In fact, model
#4 has shown an increase in resistance when fitted with a Hull Vane® . It
is expected that the results of this research would hopefully lay the ground
work for further studies and comparative analysis. The conclusions would be
beneficial for anyone interested in the Hull Vane® research.
Thesis (M.S.) - Florida Institute of Technology, 2017
2017-05-01T00:00:00ZClassification of infrasound events using hermite polynomial preprocessing and radial basis function neural networksLowrie, Christopher G.Ham, Fredric M.http://hdl.handle.net/11141/14472017-05-23T07:00:29Z2006-04-17T00:00:00ZClassification of infrasound events using hermite polynomial preprocessing and radial basis function neural networks
Lowrie, Christopher G.; Ham, Fredric M.
A method of infrasonic signal classification using hermite polynomials for signal preprocessing is presented. Infrasound is a low frequency acoustic phenomenon typically in the frequency range 0.01 Hz to 10 Hz. Data collected from infrasound sensors are preprocessed using a hermite orthogonal basis inner product approach. The hermite preprocessed signals result in feature vectors that are used as input to a parallel bank of radial basis function neural networks (RBFNN) for classification. The spread and threshold values for each of the RBFNN are then optimized. Robustness of this classification method is tested by introducing unknown events outside the training set and counting errors. The hermite preprocessing method is shown to have superior performance compared to a standard cepstral preprocessing method.
2006-04-17T00:00:00ZInvestigations on the development of a mixed displacement-pressure formulation for an anelastic displacement-field finite elementRusovici, Razvanhttp://hdl.handle.net/11141/14462017-05-23T07:00:31Z2006-03-17T00:00:00ZInvestigations on the development of a mixed displacement-pressure formulation for an anelastic displacement-field finite element
Rusovici, Razvan
Space and weapon delivery systems contain guidance components and payload that need to be protected from the extremely harsh acoustic excitation present during launch operations. The above example represents just one application where high-damping viscoelastic materials are used in the design of shock and vibration isolation components. The shock transients generally encountered are characterized by a broad frequency spectrum. Widely available finite element codes do not offer the proper tools to model the frequency-dependent mechanical properties of viscoelastic materials over the frequency domain of interest. An added difficulty is the large Poissson's ratio exhibited by some of these materials, which indicates that previously developed displacement-based finite element formulations should be complemented with mixed pressure-displacement finite element formulations. A pure displacement-based finite element generally predicts the displacements well, if the mesh used is fine enough, but the same thing may not be said about the values of the predicted stresses. The Anelastic Displacement Fields (ADF) method is employed herein to model frequency-dependence of material properties within a time-domain finite element framework and using a mixed displacement-pressure finite element formulation. Finite elements based on this new formulation are developed.
2006-03-17T00:00:00Z