.#htkoview.tex.1.3

该压缩包为最新版htk的源代码,htk是现在比较流行的语音处理软件,请有兴趣的朋友下载使用

%/* ----------------------------------------------------------- */
%/*                                                             */
%/*                          ___                                */
%/*                       |_| | |_/   SPEECH                    */
%/*                       | | | | \   RECOGNITION               */
%/*                       =========   SOFTWARE                  */ 
%/*                                                             */
%/*                                                             */
%/* ----------------------------------------------------------- */
%/* developed at:                                               */
%/*                                                             */
%/*      Speech Vision and Robotics group                       */
%/*      Cambridge University Engineering Department            */
%/*      http://svr-www.eng.cam.ac.uk/                          */
%/*                                                             */
%/*      Entropic Cambridge Research Laboratory                 */
%/*      (now part of Microsoft)                                */
%/*                                                             */
%/* ----------------------------------------------------------- */
%/*         Copyright: Microsoft Corporation                    */
%/*          1995-2000 Redmond, Washington USA                  */
%/*                    http://www.microsoft.com                 */
%/*                                                             */
%/*          2001-2002 Cambridge University                     */
%/*                    Engineering Department                   */
%/*                                                             */
%/*   Use of this software is governed by a License Agreement   */
%/*    ** See the file License for the Conditions of Use  **    */
%/*    **     This banner notice must not be removed      **    */
%/*                                                             */
%/* ----------------------------------------------------------- */
%
% HTKBook - Steve Young 1/12/97
%

\mychap{An Overview of the \HTK\ Toolkit}{htkoview}

\sidepic{toolkit}{50}{}
The basic principles of HMM-based recognition were outlined in
the previous chapter and a number of the key \HTK\ tools have already
been mentioned.  This chapter describes the software architecture
of a \HTK\ tool.   It then gives a brief outline of all the
\HTK\ tools and the way that they are used together to construct
and test HMM-based recognisers.  For the benefit of existing \HTK\ users,
the major changes in recent versions of \HTK\ are listed.
The following chapter will then illustrate
the use of the \HTK\ toolkit 
by working through a practical example of building a simple 
continuous speech recognition system.

\mysect{\HTK\ Software Architecture}{softarch}

Much of the functionality of \HTK\ is built into the library modules. 
These modules ensure that every tool interfaces to the outside world 
in exactly the same way.  They also provide a central resource of 
commonly used functions.  Fig.~\href{f:softarch} illustrates 
the software\index{software architecture}
structure of a typical \HTK\ tool and shows its input/output interfaces.

User input/output and interaction with the operating system is 
controlled by the library 
module \htool{HShell}\index{hshell@\htool{HShell}} and all memory management
is controlled by \htool{HMem}\index{hmem@\htool{HMem}}.  
Math support is provided by \htool{HMath}\index{hmath@\htool{HMath}}
and the signal processing operations needed for speech analysis are
in \htool{HSigP}\index{hsigp@\htool{HSigP}}.\index{library modules}
Each of the file types required by \HTK\ has a dedicated
interface module.  
\htool{HLabel}\index{hlabel@\htool{HLabel}} provides the interface for label files,
\htool{HLM}\index{hlm@\htool{HLM}} for language model files,
\htool{HNet}\index{hnet@\htool{HNet}} for networks and lattices,
\htool{HDict}\index{hdict@\htool{HDict}} for dictionaries,
\htool{HVQ}\index{hvq@\htool{HVQ}} for VQ codebooks and
\htool{HModel}\index{hmodel@\htool{HModel}} for HMM definitions.

\sidefig{softarch}{75}{Software Architecture}{-4}{
All speech input and output at the waveform level
is via \htool{HWave} 
and at the parameterised level via \htool{HParm}.
As well as providing a consistent interface, \htool{HWave} and 
\htool{HLabel}
support multiple file formats allowing data to be imported
from other systems.   Direct audio input is 
supported by \htool{HAudio}
and simple interactive graphics is 
provided by \htool{HGraf}.  
\htool{HUtil} provides a 
number of utility routines for manipulating
HMMs while \htool{HTrain} and \htool{HFB} contain support for the 
various \HTK\ training tools.   \htool{HAdapt} provides support
for the various \HTK\ adaptation tools.
Finally, \htool{HRec} contains 
the main recognition processing functions.
}\index{haudio@\htool{HAudio}}
\index{hrec@\htool{HRec}}
\index{hutil@\htool{HUtil}}
\index{hwave@\htool{HWave}}
\index{hparm@\htool{HParm}}
\index{hgraf@\htool{HGraf}}
\index{htrain@\htool{HTrain}}

As noted in the next section, fine control over 
the behaviour of these library modules
is provided by setting configuration variables\index{configuration variables}.  Detailed descriptions
of the functions provided by the library modules are given in the second
part of this book and the relevant configuration variables are described
as they arise.  For reference purposes, a complete list is given in
chapter~\ref{c:confvars}.

\mysect{Generic Properties of a HTK Tool}{genprops}

\HTK\ tools are designed to run with a traditional command-line style interface.
Each tool\index{command line!options}
has a number of required arguments plus optional arguments.
The latter are always prefixed by a minus sign.  As an example,
the following command would invoke the mythical \HTK\ tool called
\htool{HFoo}
\begin{verbatim}
    HFoo -T 1 -f 34.3 -a -s myfile file1 file2
\end{verbatim}
This tool has two main arguments called \texttt{file1} and
\texttt{file2} plus four optional arguments.  Options
are always introduced by a single letter option name followed
where appropriate by the option value.  The option value
is always separated from the option name by a space. Thus, the value of the
\texttt{-f} option is a real number, the value of the
\texttt{-T} option is an integer number and the value of the
\texttt{-s} option is a string.  The \texttt{-a} option has no following
value and it is used as a simple flag to enable or disable some
feature of the tool.  Options whose names are a capital letter
have the same meaning across all tools.  For example, the \texttt{-T}
option is always used to control the trace output of a \HTK\ tool.

In addition to command line arguments, the operation of a tool
can be controlled by parameters stored 
in a configuration file\index{configuration files}.
For example, if the command 
\begin{verbatim}
    HFoo -C config -f 34.3 -a -s myfile file1 file2
\end{verbatim}
is executed, the tool \htool{HFoo} will load the 
parameters stored in the configuration
file \texttt{config} during its initialisation procedures.  Multiple
configuration files can be specified by repeating the \verb|-C|
option, e.g.\ 
\begin{verbatim}
    HFoo -C config1 -C config2 -f 34.3 -a -s myfile file1 file2
\end{verbatim}
Configuration parameters can sometimes be used as an
alternative to using command line arguments.  For example, trace
options can always be set within a configuration file.  However, the
main use of configuration files is to control the detailed behaviour
of the library modules on which all \HTK\ tools depend.

Although this style of command-line 
working may seem old-fashioned when compared to modern
graphical user interfaces, it has many advantages.  In particular,
it makes it simple to write shell scripts to control \HTK\ tool execution.  This
is vital for performing large-scale system building and experimentation.
Furthermore, defining all operations using text-based commands allows
the details of system construction or experimental procedure to
be recorded and documented.

Finally, note that a summary of the command line and
options for any \HTK\ tool can be obtained simply by executing
the tool with no arguments.

\mysect{The Toolkit}{toolkit}

The \HTK\ tools are best introduced by going through the
processing steps involved in building a sub-word based continuous speech 
recogniser.  As shown in Fig.~\href{f:sysoview}, there are 4
main phases: data preparation, training, testing and analysis.

\subsection{Data Preparation Tools}

\index{data preparation}
In order to build a set of HMMs, a set of speech data files
and their associated transcriptions are required.  Very often
speech data will be obtained from database archives, typically
on CD-ROMs.  Before it can be used in training, it must be 
converted into the appropriate parametric form and any associated
transcriptions must be converted to have the correct format
and use the required phone or word labels.  If the speech needs to be
recorded, then the tool \htool{HSLab}\index{hslab@\htool{HSLab}} can be used both to record the
speech and to manually annotate it with any required transcriptions.

Although all \HTK\ tools can parameterise waveforms \textit{on-the-fly}, in practice
it is usually better to
parameterise the data just once.  The tool \htool{HCopy}\index{hcopy@\htool{HCopy}}
is used for this.  As the name suggests, \htool{HCopy} is used to copy one
or more source files to an output file.  
Normally, \htool{HCopy} copies the whole file, but a variety
of mechanisms are provided for extracting segments of files and concatenating
files.  By setting the appropriate configuration variables, all input files
can be converted to parametric form as they are read-in.
Thus, simply copying each file in this manner performs the required encoding.
The tool \htool{HList}\index{hlist@\htool{HList}} can be used to check the contents of any speech file
and since it can also convert input on-the-fly, it can be used to check
the results of any conversions before processing large quantities of data.
Transcriptions will also need preparing.  Typically the labels used in the
original source transcriptions will not be exactly as required, for example,
because of differences in the phone sets used.  Also, HMM training might
require the labels to be context-dependent.  The tool \htool{HLEd}\index{hled@\htool{HLEd}} is
a script-driven label editor which is designed to make the required transformations
to label files.  \htool{HLEd} can also output files  to a single \textit{Master Label
File} MLF which is usually more convenient for subsequent processing.
Finally on data preparation, \htool{HLStats}\index{hlstats@\htool{HLStats}} can gather and display statistics
on label files and where required, \htool{HQuant}\index{hquant@\htool{HQuant}} can be used to build a
VQ codebook in preparation for building  discrete probability HMM system.

\subsection{Training Tools}

The second step of system building is to\index{training tools}
define the topology required for each HMM by writing a prototype definition.
\HTK\ allows HMMs to be built with any desired topology.
HMM definitions can be stored externally as simple text files and
hence it is possible to edit them with any convenient text
editor. Alternatively, the standard \HTK\ distribution includes
a number of example HMM prototypes and a script to generate
the most common topologies automatically.
With the exception of the transition
probabilities, all of the HMM parameters given in 
the prototype definition\index{prototype definition}
are ignored.  The purpose of the prototype definition is only
to specify the overall characteristics and topology of the HMM.  The
actual parameters will be computed later by the training tools.  
Sensible values for
the transition probabilities must be given but the training
process is very insensitive to these.  An acceptable and  simple strategy
for choosing these probabilities is to make all of the transitions
out of any state equally likely.


\centrefig{sysoview}{100}{\HTK\ Processing Stages}

The actual training process takes place in stages and it is
illustrated in more detail in Fig.~\href{f:tsubword}.  
Firstly, an initial set of models must be created.  If there is
some speech data available for which the location of the sub-word (i.e.\ phone)
boundaries have been marked, then this can be used as \textit{bootstrap data}.
In this case, the tools \htool{HInit}\index{hinit@\htool{HInit}} 
and \htool{HRest}\index{hrest@\htool{HRest}}
provide {\it isolated word} style
training using the fully labelled bootstrap\index{bootstrapping} data.  Each of the required
HMMs is generated individually.  \htool{HInit} reads in all of the bootstrap
training data and {\it cuts out} all of the examples of the required
phone.  It then iteratively computes an initial set of parameter values
using a {\it segmental k-means} procedure\index{segmental k-means}.  On the first cycle, the training data
is uniformly segmented, each model state is matched with the
corresponding data segments and then means and variances are estimated.
If mixture Gaussian models are being trained, then a modified form
of k-means clustering is used.  On the second and successive cycles,
the uniform segmentation is replaced by Viterbi alignment.  
The initial parameter values computed by \htool{HInit} are then further re-estimated
by \htool{HRest}.  Again, the fully labelled bootstrap data is used but this
time the segmental k-means procedure is replaced by the Baum-Welch re-estimation
procedure described in the previous chapter.  When no bootstrap data is
available, a so-called \textit{flat start} can be used.  In this case all
of the phone models are initialised to be identical and have state means
and variances equal to the global speech mean and variance.  The tool
\htool{HCompV}\index{hcompv@\htool{HCompV}} can be used for this.\index{flat start}

\centrefig{tsubword}{90}{Training Sub-word HMMs}

Once an initial set of models has been created, the tool \htool{HERest}
is used to perform {\em embedded training} using the 
entire\index{embedded training}
training set.  \htool{HERest}\index{herest@\htool{HERest}} performs a single Baum-Welch
re-estimation of the whole set of HMM phone models simultaneously.  For each
training utterance, the corresponding phone models are concatenated and then
the forward-backward algorithm is used to accumulate the statistics of state
occupation, means, variances, etc., for each HMM in the sequence.  When
all of the training data has been processed, the accumulated statistics
are used to compute re-estimates of the HMM parameters.  \htool{HERest} is
the core \HTK\ training tool.  It is designed to process large databases, it has
facilities for pruning\index{pruning} to reduce computation and it can be run in parallel 
across a network of machines.