Stellenbosch Electrical and Electronic Engineering report, thesis and dissertation (H. Kamper)

\chapter{Project Planning Schedule}
\makeatletter\@mkboth{}{Appendix}\makeatother
\label{appen:derivations_bigramseg}

This is an appendix.

\chapter{Outcomes Compliance}
\makeatletter\@mkboth{}{Appendix}\makeatother
\label{appen:derivations_bigramseg}

This is another appendix.

\graphicspath{{conclusion/fig/}}

\chapter{Summary and Conclusion}
\label{chap:conclusion}

\chapter*{Abstract}
\addcontentsline{toc}{chapter}{Abstract}
\makeatletter\@mkboth{}{Abstract}\makeatother

The English abstract.



\selectlanguage{afrikaans}
\chapter*{Opsomming}
\addcontentsline{toc}{chapter}{Opsomming}
\makeatletter\@mkboth{}{Opsomming}\makeatother

Die Afrikaanse uittreksel.

\selectlanguage{english}

\chapter*{Acknowledgements}
% \addcontentsline{toc}{chapter}{Acknowledgements}
\makeatletter\@mkboth{}{Acknowledgements}\makeatother

I would like to thank my dog, Muffin. I also would like to thank the inventor of the incubator; without him/her, I would not be here. Finally, I would like to thank Dr Herman Kamper for this amazing report template.

%\chapter*{Declaration}
\newpage
\thispagestyle{plain}
\addcontentsline{toc}{chapter}{Declaration}
\makeatletter\@mkboth{}{Declaration}\makeatother

\vspace*{-25pt}

\section*{\centering Plagiaatverklaring / \textit{Plagiarism Declaration}}

\vspace*{5pt}

\begin{enumerate}
    \item Plagiaat is die oorneem en gebruik van die idees, materiaal en ander intellektuele eiendom van ander persone asof dit jou eie werk is.\\
    \textit{Plagiarism is the use of ideas, material and other intellectual property of another's work
        and to present is as my own.}
    
    \item Ek erken dat die pleeg van plagiaat 'n strafbare oortreding is aangesien dit 'n vorm van diefstal is.\\
    \textit{I agree that plagiarism is a punishable offence because it constitutes theft.}
    
    \item Ek verstaan ook dat direkte vertalings plagiaat is. \\
    \textit{I also understand that direct translations are plagiarism.}
    
    \item Dienooreenkomstig is alle aanhalings en bydraes vanuit enige bron (ingesluit die internet) volledig verwys (erken). Ek erken dat die woordelikse aanhaal van teks sonder aanhalingstekens (selfs al word die bron volledig erken) plagiaat is. \\
    \textit{Accordingly all quotations and contributions from any source whatsoever (including the internet) have been cited fully. I understand that the reproduction of text without quotation marks (even when the source is cited) is plagiarism}
    
    \item Ek verklaar dat die werk in hierdie skryfstuk vervat, behalwe waar anders aangedui, my eie oorspronklike werk is en dat ek dit nie vantevore in die geheel of gedeeltelik ingehandig het vir bepunting in hierdie module/werkstuk of 'n ander module/werkstuk~nie. \\
    \textit{I declare that the work contained in this assignment, except where otherwise stated, is my original work and that I have not previously (in its entirety or in part) submitted it for grading in this module/assignment or another module/assignment.}
\end{enumerate}

\vfill

\noindent \begin{tabularx}{1.0\linewidth}{|L|L|}
    \hline
	& \\[-10pt]
    \textbf{24702811} & \includegraphics[height=1.5cm]{frontmatter/fig/benjamin_franklin_signature.png} \\
%    \vspace{1cm}
    {Studentenommer / \textit{Student number}} & {Handtekening / \textit{Signature}} \\
    \hline
    & \\[-10pt]
    \textbf{A.\ Daneels} & \textbf{\today} \\
    {Voorletters en van / \textit{Initials and surname}} & {Datum / \textit{Date}} \\
    \hline
\end{tabularx}

\vspace{15pt}

\chapter*{Nomenclature\markboth{}{Nomenclature}}
\addcontentsline{toc}{chapter}{Nomenclature}

% \vspace*{-3mm}
\subsubsection*{Variables and functions}

\begingroup
\renewcommand{\arraystretch}{1.2}
\renewcommand{\tabularxcolumn}[1]{p{#1}}
\begin{tabularx}{\textwidth}{@{}p{2.5cm}L}
    $p(x)$ & Probability density function with respect to variable $x$.\\
    $P(A)$ & Probability of event $A$ occurring.\\
    $\varepsilon$ & The Bayes error. \\
    $\varepsilon_u$ & The Bhattacharyya bound. \\
    $B$ & The Bhattacharyya distance. \\
    $s$ & An HMM state.  A subscript is used to refer to a particular state, e.g.\ $s_i$ refers to the $i^{\text{th}}$ state of an HMM. \\
    $\mathbf{S}$ & A set of HMM states. \\
    $\mathbf{F}$ & A set of frames. \\
    $\mathbf{o}_f$ & Observation (feature) vector associated with frame $f$. \\
    $\gamma_s(\mathbf{o}_f)$ & A posteriori probability of the observation vector $\mathbf{o}_f$ being generated by HMM state $s$. \\
    $\mu$ & Statistical mean vector. \\
    $\Sigma$ & Statistical covariance matrix. \\
    $L(\mathbf{S})$ & Log likelihood of the set of HMM states $\mathbf{S}$ generating the training set observation vectors assigned to the states in that set. \\
    $\mathcal{N}(\mathbf{x} | \mu, \Sigma)$ & Multivariate Gaussian PDF with mean $\mu$ and covariance matrix $\Sigma$.\\
    $a_{ij}$ & The probability of a transition from HMM state $s_i$ to state $s_j$. \\
    $N$ & Total number of frames or number of tokens, depending on the context. \\
    $D$ & Number of deletion errors. \\
    $I$ & Number of insertion errors. \\
    $S$ & Number of substitution errors. \\
\end{tabularx}
\endgroup


\newpage
\subsubsection*{Acronyms and abbreviations}

\begingroup
\renewcommand{\arraystretch}{1.2}
\begin{tabular}{@{}p{2.5cm} l}
    AE      & Afrikaans English \\
    AID     & accent identification \\
    ASR     & automatic speech recognition \\
    AST     & African Speech Technology \\
    CE      & Cape Flats English \\
    DCD     & dialect-context-dependent \\
    DNN		& deep neural network \\
    G2P     & grapheme-to-phoneme \\
    GMM     & Gaussian mixture model \\
    HMM     & hidden Markov model \\
    HTK     & Hidden Markov Model Toolkit \\
    IE      & Indian South African English \\
    IPA     & International Phonetic Alphabet \\
    LM      & language model \\
    LMS     & language model scaling factor \\
    MFCC    & Mel-frequency cepstral coefficient \\
    MLLR    & maximum likelihood linear regression \\
    OOV     & out-of-vocabulary \\
    PD      & pronunciation dictionary \\
    PDF     & probability density function \\
    SAE     & South African English \\
    SAMPA   & Speech Assessment Methods Phonetic Alphabet \\
\end{tabular}
\endgroup

\graphicspath{{frontmatter/fig/}}
\pagenumbering{Alph}

\begin{titlepage}

	\begin{center}			
		\vfill
        \vfill
        \vfill
		{\bfseries \huge A critical analysis of design flaws in the Death Star \par}
		\vfill
        {\large by \\[5pt]}
		{\Large {\Large Luke Skywalker} \par}
		\vfill
		\vfill

		% % Skripsie
		% {\large Report presented in partial fulfilment of the requirements of the module \\ Project (E) 448 for the degree Baccalaureus in Engineering (Electrical and Electronic) in the Faculty of Engineering at Stellenbosch University. \par}        

		% Masters (Research)
		{\large Thesis presented in fulfilment of the requirements for the degree of \\ Master of Engineering (Electronic) in the Faculty of Engineering at Stellenbosch University \par}
		
		% % Masters (Structured)
		% {\large Research assignment presented in partial fulfilment of the requirements for the degree of Master of Engineering (Electronic) \\ in the Faculty of Engineering at Stellenbosch University \par}
		
		% % PhD
		% {\large Dissertation presented for the degree of Doctor of Philosophy (Electronic Engineering) in the Faculty of Engineering at Stellenbosch University \par}

		\vfill
		
		{\large {Supervisor}: Dr Obi W.\ Kenobi}\par
    	{\large {Co-supervisor}: Prof Minch Yoda}

		\vfill
		{\large December 2099}
        \vfill
	\end{center}
\end{titlepage}

\graphicspath{{introduction/fig/}}

\chapter{Introduction}
\label{chap:introduction}

The last few years have seen great advances in speech recognition. Much of this progress is due to the resurgence of neural networks; most speech systems now rely on deep neural networks (DNNs) with millions of parameters~\cite{dahl+etal_taslp12,hinton+etal_spm2012}.
However, as the complexity of these models has grown, so has their reliance on labelled training data. Currently, system development requires large corpora of transcribed speech audio data, texts for language modelling, and pronunciation dictionaries.
Despite speech applications becoming available in more languages, it is hard to imagine that resource collection at the required scale would be possible for all 7000 languages spoken in the world today.

I really like apples.

\section{Section heading}

This is some section with two table in it: Table~\ref{tbl:exemplars} and Table~\ref{tbl:abx_speaker}.

\begin{table}[!h]
    \mytable
    \caption{Performance of the unconstrained segmental Bayesian model on TIDigits1 over iterations in which the reference set is refined.}
    \begin{tabularx}{\linewidth}{@{}lCCCCC@{}}
        \toprule
        Metric     & 1 & 2 & 3 & 4 & 5 \\
        \midrule
        WER (\%)                        & $35.4$ & $23.5$ & $21.5$ & $21.2$ & $22.9$ \\
        Average cluster purity (\%)       & $86.5$ & $89.7$ & $89.2$ & $88.5$ & $86.6$ \\
        Word boundary $F$-score (\%)         & $70.6$ & $72.2$ & $71.8$ & $70.9$ & $69.4$ \\
        Clusters covering 90\% of data   & 20             & 13 & 13 & 13 & 13 \\
        \bottomrule
    \end{tabularx}
    \label{tbl:exemplars}
\end{table}


\begin{table}[!h]
    \renewcommand{\arraystretch}{1.1}
    \centering
    \caption{A table with an example of using multiple columns.}
    \begin{tabularx}{0.65\linewidth}{@{}lCCr@{}}
        \toprule
        & \multicolumn{2}{c}{Accuracy (\%)} \\
        \cmidrule(lr){2-3}
        Model    & Intermediate & Output & Bitrate\\
        \midrule
        Baseline & 27.5         & 26.4   & 116 \\
        VQ-VAE   & 26.0         & 22.1   & 190 \\
        CatVAE   & 28.7         & 24.3   & 215 \\
        \bottomrule
    \end{tabularx}
    \label{tbl:abx_speaker}
\end{table}

\newpage

This is a new page, showing what the page headings looks like, and showing how to refer to a figure like Figure~\ref{fig:cae_siamese}.

\begin{figure}[!t]
    \centering
%     \includegraphics[width=\linewidth]{cae_siamese}
    \includegraphics[width=0.918\linewidth]{cae_siamese}
    \caption[I am the short caption that appears in the list of figures, without references.]{
    (a) The cAE as used in this chapter. The encoding layer (blue) is chosen based on performance on a development set.
    (b) The cAE with symmetrical tied weights. The encoding from the middle layer (blue) is always used.
    (c) The siamese DNN. The cosine distance between aligned frames (green and red) is either minimized or maximized depending on whether the frames belong to the same (discovered) word or not.
    A cAE can be seen as a type of DNN~\cite{dahl+etal_taslp12}.
    }
    \label{fig:cae_siamese}
\end{figure}


The following is an example of an equation:
\begin{equation}
P(\vec{z} | \vec{\alpha}) = \int_{\vec{\pi}} P(\vec{z} | \vec{\pi}) \, p(\vec{\pi} | \vec{\alpha}) \, \textrm{d} \vec{\pi}
= \int_{\vec{\pi}} \prod_{k = 1}^K \pi_k^{N_k} \frac{1}{B(\vec{\alpha})} \prod_{k = 1}^K \pi_k^{\alpha_k - 1} \, \textrm{d} \vec{\pi}
\label{eq:example_equation}
\end{equation}
which you can subsequently refer to as~\eqref{eq:example_equation} or Equation~\ref{eq:example_equation}.
But make sure to consistently use the one or the other (and not mix the two ways of referring to equations).

\section{Contributions}

The following papers resulted from the work presented here: \\

\begin{tcolorbox}[width=\linewidth, colback=white!95!black, boxrule=0.5pt]
	\small
	\textit{Conference paper 1:} \\
	L. Skywalker, D. Vadar, and O. W. Kenobi, ``A comparison between father-son and master-apprentice relationships in space conflict situations,'' in \textit{Proceedings of the International Conference on Action, Space and Star Politics (ICASSP)}, 2020. \\
	
	\textit{Journal paper 1:} \\
	L. Skywalker, and L. Organa, ``Identifying weaknesses in large evil corporations,'' \textit{IEEE Transactions on the Exploration of the Outer Rim}, vol. 21, pp. 154--174, 2021.	
\end{tcolorbox}

MIT License

Copyright (c) 2020-2024 Herman Kamper

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

% This file was created with JabRef 2.10.
% Encoding: ISO8859_1


@Article{dahl+etal_taslp12,
  Title                    = {Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition},
  Author                   = {Dahl, George E and Yu, Dong and Deng, Li and Acero, Alex},
  Journal                  = {IEEE Trans. Audio, Speech, Language Process.},
  Year                     = {2012},
  Number                   = {1},
  Pages                    = {30--42},
  Volume                   = {20},

  Owner                    = {kamperh},
  Quality                  = {1},
  Timestamp                = {2014.09.15}
}

@Article{hinton+etal_spm2012,
  Title                    = {Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups},
  Author                   = {Hinton, Geoffrey and Deng, Li and Yu, Dong and Dahl, George E and Mohamed, Abdel-Rahman and Jaitly, Navdeep and Senior, Andrew and Vanhoucke, Vincent and Nguyen, Patrick and Sainath, Tara N and Kingsbury, B},
  Journal                  = {{IEEE} Signal Process. Mag.},
  Year                     = {2012},
  Number                   = {6},
  Pages                    = {82--97},
  Volume                   = {29},

  Owner                    = {kamperh},
  Quality                  = {1},
  Timestamp                = {2014.09.15}
}

\documentclass[a4paper,12pt]{report}

% Page layout
\usepackage[left=2.5cm,right=2.5cm,top=2.5cm,bottom=2.5cm]{geometry}

% Font and text
\usepackage[afrikaans,english]{babel}
\usepackage{microtype}
\usepackage{setspace}
\usepackage{lmodern}
\usepackage{siunitx}
\usepackage{tcolorbox}
% \usepackage[scaled=.96]{XCharter}
\usepackage[scaled=.96]{caladea}  % font required by Stellenbosch University
\usepackage[scaled=.96,lf]{carlito}  % the scaling makes the math the same height
\renewcommand*\oldstylenums[1]{\carlitoOsF #1}
\newcommand{\myemph}[1]{{\sffamily\bfseries#1}}
\sloppy
\onehalfspacing

% Headings
\usepackage[raggedright,sf,bf]{titlesec}
\usepackage[margin=\the\parindent,small,bf,sf]{caption}
\titlelabel{\thetitle.\ }
\titleformat{\chapter}[display]{\huge\bfseries\sffamily}{\chaptertitlename\ \thechapter}{15pt}{\Huge \raggedright}
\titlespacing*{\chapter}{0pt}{0pt}{40pt}  % remove spacing before chapter headings
\makeatletter
\let\originall@chapter\l@chapter
\def\l@chapter#1#2{\originall@chapter{{\sffamily #1}}{#2}}
\makeatother

%% Alternative headings using small-caps (comment out the top section)
%\usepackage[raggedright,bf]{titlesec}
%\usepackage[margin=\the\parindent,small,bf]{caption}
%\titlelabel{\thetitle.\ }
%\titleformat{\chapter}[display]{\huge\scshape}{\chaptertitlename\ \thechapter}{15pt}{\Huge \raggedright}
%\titlespacing*{\chapter}{0pt}{0pt}{40pt}  % remove spacing before chapter headings

% Table of contents
\let \savenumberline \numberline
\def \numberline#1{\savenumberline{#1.}}

% Figures
\usepackage{graphicx}
\usepackage{tikz}
\usepackage{pdfpages}
\usepackage{subcaption}
\setlength{\abovecaptionskip}{7.5pt}  % spacing above and below captions
\newcommand*{\WaterMark}[2][0.2\paperwidth]{\AddToShipoutPicture*{\AtTextCenter{\parbox[c]{0pt}{\makebox[0pt][c]{\includegraphics[width=#1]{#2}}}}}}

% Mathematics
\usepackage[cmex10]{amsmath}
\usepackage{amssymb}
\usepackage{cancel}
\DeclareMathOperator*{\argmax}{arg\,max}
\newcommand{\T}{^\top}
\newcommand{\tr}{\textrm{tr}}
\renewcommand{\vec}[1]{\boldsymbol{\mathbf{#1}}}
\newcommand{\defeq}{\triangleq}

% Tables
\usepackage{booktabs}
\usepackage{tabularx}
\usepackage{multirow}
\newcommand{\mytable}{
    \centering
    \small
    \renewcommand{\arraystretch}{1.2}
    }
\renewcommand{\tabularxcolumn}[1]{m{#1}}
\newcolumntype{C}{>{\centering\arraybackslash}X}
\newcolumntype{L}{>{\raggedright\arraybackslash}X}

% Header and footer
\usepackage{fancyhdr}
\pagestyle{fancy}
\fancyhf{}
\renewcommand{\sectionmark}[1]{\markright{\normalsize \thesection.\ #1}}
\fancyhead[C]{\nouppercase{\textit{\rightmark}}}
\fancyhead[RO]{\thepage}
 \fancyhead[LE]{\thepage}  % double-sided printing
\fancyfoot{}
\setlength\headheight{14.5pt}
\renewcommand{\headrulewidth}{0pt}
\fancypagestyle{plain}{\fancyhead{}
                       \renewcommand{\headrulewidth}{0pt}
                       \fancyfoot[C]{\thepage}}

% Pseudo-code
\usepackage{algorithm}  % should go before \usepackage{hyperref}

% Table of contents and hyperlinks
\usepackage{hyperref}
\hypersetup{colorlinks=true,linktoc=all,citecolor=black,linkcolor=black}
\usepackage[nottoc]{tocbibind}

% Pseudo-code
\usepackage{algpseudocode}  % should go after \usepackage{hyperref}
\renewcommand{\thealgorithm}{\arabic{chapter}.\arabic{algorithm}} 
\captionsetup[algorithm]{labelfont={bf,sf},font=small,labelsep=colon}

% Bibliography
\usepackage{cite}  % automatically reorder inline citations
\bibliographystyle{IEEEtran}

% Fix titlesec issue
\usepackage{etoolbox}
\makeatletter
\patchcmd{\ttlh@hang}{\parindent\z@}{\parindent\z@\leavevmode}{}{}
\patchcmd{\ttlh@hang}{\noindent}{}{}{}
\makeatother


\begin{document}

% Front matter
\input{frontmatter/title_page}
\pagenumbering{roman}
\input{frontmatter/declaration}
\input{frontmatter/abstract}
\input{frontmatter/acknowledgements}
\tableofcontents
%\listoffigures
%\listoftables
\input{frontmatter/nomenclature}
\newpage
\pagenumbering{arabic}

% Contents
\input{introduction/introduction}
\input{conclusion/conclusion}

% Bibliography
\bibliography{mybib}

% End matter
\appendix
\input{appendices/appendices}

\end{document}