\tikzset{% >={Latex[width=2mm,length=2mm]}, base/.style = {rectangle, rounded corners, draw=black, minimum width=4cm, minimum height=1cm, text centered, font=\sffamily}, recorder/.style = {base, fill=blue!30}, meta/.style = {base, fill=red!30}, annotator/.style = {base, fill=green!30}, txt/.style = {font=\sffamily, text centered}, title/.style = {txt, font=\sffamily\Large}, } % Drawing part, node distance is 1.5 cm and every node % is prefilled with white background \begin{tikzpicture}[node distance=5cm, every node/.style={fill=white, font=\sffamily}, align=center] % Specification of nodes (position, etc.) % \node (lena) [recorder] {LENA recorder}; % \node (babylogger) [recorder, right of=lena, xshift=5em] {BabyLogger}; % \node (others) [recorder, right of=babylogger] {Other alternatives\\ % \footnotesize{USB, Olympus...}}; % \node (lena_software) [classifier, below of=lena, yshift = 5em] {LENA software\\ % \footnotesize{Speaker type, Adult Word Count,}\\ % \footnotesize{Child Vocalization Count, Conversational Turn Count}}; % \node [txt, below of=lena_software, yshift=10em] {\large{LENA commercial environment}}; % \node (vtc) [classifier, below of=babylogger, xshift=8em, yshift = 7em] {Voice Type Classifier (VTC)\\ % \footnotesize {speech detection, speaker type classification}}; % \node (alice) [classifier, below of=vtc, yshift = 8em] {Automatic LInguistic Unit Count Estimator (ALICE)\\ % \footnotesize {phoneme, syllable and word counts}}; % \node (seshat) [annotator, below of=lena_software, xshift=15em, yshift = 2em] {Seshat\\ % \footnotesize{web-based annotator}\\ % \footnotesize{inter-rater reliability}}; % \node (zooniverse) [annotator,right of=seshat] {Zooniverse\\ % \footnotesize{crowd-sourced classification tasks}}; % \node (elan) [annotator, left of=seshat] {ELAN\\ % \footnotesize{annotation software}}; % \node (das) [annotator, below of=elan, yshift = 10em] {ACLEW DAS\\ % \footnotesize{annotation scheme}}; % \node (recorders) [title,right of=others] {Recording device}; % \node (classifiers) [title,below of=recorders, yshift=3em] {Automatic annotation}; % \node (annotators) [title,below of=classifiers, yshift=3em] {Manual annotation}; % % Specification of lines between nodes specified above % % with aditional nodes for description % \draw[->] (lena) -- (lena_software); % \draw[->] (vtc) -- (alice); % \draw[->] (elan) -- (das); % \draw [draw=black,dashed] ($(lena.north west) + (-2,0.5)$) rectangle ($(lena_software.south east) + (1,-1.5)$); \node (media) [recorder] { \Large \textbf{Media} \normalsize \\ ($\sim 10^2$ to $10^4$ hours) \\ \framebox{ {\begin{varwidth}{\linewidth}\begin{itemize} \item Audio (up to 24 hours \\ per recording) \item Video (up to 30 minutes \\ per recording) \item Accelerometer data (xyz) \item etc. \end{itemize}\end{varwidth}} }}; \node (annotations) [annotator,shape=rectangle,draw,right of=media] { \Large \textbf{Annotations} \normalsize \\ ($\sim 10^5$ to $10^7$ segments)\\ \\ \begin{tabular}{l r} \textbf{manual} & \textbf{automated}\\ \multicolumn{2}{c}{who-speaks-when} \\ \multicolumn{2}{c}{linguistic units} \\ \multicolumn{2}{c}{vocal maturity} \\ speech directedness & \\ transcriptions & \\ \end{tabular} }; \node (metadata) at ($(media)!0.5!(annotations)-(0,3)$) [meta,shape=rectangle,draw] { \Large \textbf{Metadata} \normalsize \\ \\ \framebox{ {\begin{varwidth}{\linewidth}\begin{itemize} \item Recordings date and time, type of device, etc. \item Child date of birth, gender, normativity, etc. \item Socio-economic status, location, language(s), household size, etc. \item Questionnaires \end{itemize}\end{varwidth}} }}; \end{tikzpicture}