RESPITE CASA Toolkit CTK v1 . 1 . 1 User's Guide Jon Barker
Contents
1. 2K 2K 2K a 2K a a K ara ak ak K K Tutorial 1 A spectrogram of the addition of 3 sinusoids INCLUDE CTKROOT scripts spectrogram ctk BLOCK main ADD ii SineWave DURATION i SAMPLE RATE 1000 FREQ 1 ADD i2 SineWave DURATION i SAMPLE RATE 1000 FREQ 2 ADD i3 SineWave DURATION 1 SAMPLE RATE 1000 FREQ 3 ADD a Adder NINPUTS 3 ADD 41 1 BEFORE_PLOT subplot 2 1 1 ADD s Spectrogram ADD d2 MDisplay BEFORE PLOT subplot 2 1 2 CONNECT 11 a int CONNECT i2 a in2 CONNECT i3 a in3 ENDBLOCK FE gt k KK ok ok ok k k ok ok ok ok ok ok ok ak ok ok ok ok ok ok 2 2 oko k ook ke ok ok K K 2k 2k 2k ok 2 ok 2 2 ok ok k k KK K Figure A 1 Tutorial 1 usr local bin CTKScript HE af a af aa aa a ea ak ak 3K K a a 3K 2K a II af a ae a a ak FE ak a ak ak 2K Tutorial 2 The Delay Line BLOCK tap ADD di Delay DELAY 20 ADD mi Scale X 0 5 ADD ti Tee INPUT ini d1 OUTPUT outi ti outi OUTPUT out2 t1 out2 PARAMETER SCALE m1 X PARAMETER DELAY d1 DELAY ENDBLOCK BLOCK main ADD ADD ADD ADD ADD ADD ADD ADD ii SineWave DURATION 0 5 FREQ 5 SAMPLE RATE 1000 tapi tap tap2 tap tap3 tap tap4 tap tapb tap first output MDisplay BEFORE PLOT subplot 2 1 1 second output MDisplay BEFORE_PLOT subplot 2 1 2 CONNECT tapi out2 first output CONNECT tap4 out2 second output ENDBLOCK FE gt K KK ok ok ok k ak ok ok2. Input Socket Reset Compute Compute Output Socket Figure 3 3 A complete system The CTK Graphical User Interface GUI Although writing scripts for small simple systems is straighforward more com plicated system can be hard to visualise from the linear script description A graphical interface has been provided that allows the user to design systems and examine and edit existing scripts in an intuitive manner The interface can also be used as a platform from which to interactively experiment with CTK systems Use of the CTK GUI is cover in Chapter 5 11 Extending the toolkit The toolkit is distributed with a library of existing primitive blocks These primitive blocks are written in C and compiled into the toolkit However the toolkit is designed to be easily extendible It is hoped that with the help of developer s documentation and the example of the source code for the existing inbuilt blocks users with a little knowledge of C will be able to write their own inbuilt blocks for incorporation into the toolkit library Chapter 6 describes the necessary steps for adding a new primitive block to the system Chapter 4 Writing CASA Toolkit Scripts This chapter is composed of a number of walk through tutorials which illustrate how to build simple systems using the CASA Definition Language 4 1 Tutorial 1 Displaying a spectrogram of su perimposed sinusoids In the first example a system will be buil
3. 3 are command line variables You can now execute the com mand by proceeding it with three argument values For example try typing examplei ctk 2 3 4 Or examplei ctk 100 200 300 or with any other three frequency values after the command name We now wish to make a spectrogram of the signal To do this we need to break the signal into a series of windowed frames and calculate the magnitude of the complex FFT of each frame This is done using the two more inbuilt blocks 16 named Frame and FFT connected together in series However a spectrogram is a commonly used representation and it is rather obscure and cumbersome to have to keep writing out its inbuilt block definition in full every time we wish to use it e g to construct a spectrogram we need to add something like the following to the script ADD frame Frame ADD fft FFT and it would be better is we could simply type ADD s Spectrogram We can in fact use this neater alternative if we first define an intermediate level block called Spectrogram in terms of the the inbuilt blocks Frame and FFT number of useful intermediate level block definitions including the spectrogram block come supplied with the toolkit and are stored in the toolkit scripts di rectory So rather than defining our own Spectrogram intermediate level block we can simply use the INCLUDE command to add this prewritten block to our script So directly before the line BLOCK main type INCLUDE
4. bin 6 2 Writing Your Own Primitive Block The within CTKLOCAL src there are files named ctk_NEW_BLOCK_TEMPLATE cpp and ctk_NEW_BLOCK_TEMPLATE hh These files provide an annotated template for the code you need to implement when writing your own inbuilt blocks The files my_blocks cpp and my_blocks hh contain code for a few simple example blocks Further examples are in the CTKROOT src directory More detailed CTK developers documentation should eventually be available Once you have written the cpp and hh files for the block you wish to add you need to perform the following three steps 1 Add your new hh file to the list of includes in the file block headers hh 2 Add a line to the file translation table according to the instructions it contains 3 Edit the file Makefile localin CTKLOCAL src and change the definition of OBJS to include the file you have added Once these steps have been completed move to the directory CTKROOT src and type make and the local CTK binaries should now be rebuilt to include your new block in the CTK library To check that the rebuild has been successful you can type ctk to start up the GUI and check that your block appears correctly in the pull down block menu Appendix Appendix A Scripts for User Guide Tutorials A 1 Script for Tutorial 1 A 2 Script for Tutorial 2 A 3 Script for Tutorial 3 31 32 usr local bin CTKScript HEK akak af af ak 2k aa
5. ok ok ok ok ok o ok ok ok K 2K ok ok ok oko 2 ook ke ok ok okok ok ok ok ok 2k k 2 ok oe ok ok k Figure A 2 Tutorial 2 33 34 usr local bin CTKScript FE gt k KK ok ok ok k ak ok ok ok ak 2 ak ok ak ok ok ok 2 2K ok ok 2 k ak 2 k ak ke ok ok K K ok 2k 2K k 2k k 2 2 oko ok ok k K ok KK Tutorial 3 Missing Data Speech Recognition BLOCK ADD main i BinaryInputFile FILE_NAME 1 HEADER_SIZE 256 BYTES_PER_SAMPLE 4 SWAP BYTES Yes SAMPLE RATE 100 SAMPLES PER FRAME 64 FLOATING_POINT Yes ADD ADD ADD ADD ADD ADD ADD ADD ADD teei Tee dbtoamp DBToAmp tee2 Tee nest Stationar yNoiseEst imation NFRAMES 10 floor Floor X 0 0 pow2 Power X 2 0 scale Scale X 2 0 comp Comparitor X 0 0 decoder HMMDecoderMD HMM_FILE mlist LABELS 1234567890zs SILENCE s USE_ERF_TABLE T USE_BOUNDS T WORD_PENALTY 0 TRANSCRIPTION 2 ADD powi Power 2 0 CONNECT nest out2 floor CONNECT tee2 out2 CONNECT powi comp in2 CONNECT teei outi decoder ini ENDBLOCK oko ok ok k k KK KK 2k 2k 2 2k Figure A 3 Tutorial 3
6. the CASA Toolkit graphical user interface GUI This in terface has been built on top of the toolkit to aid the design of more complicated systems The GUI allows systems to be designed in an intuitive manner by ar ranging widgets on a canvas The graphical representation can then be saved as a CTK Script that can be run in the normal way The GUI can also be used to visualise and edit existing scripts The toolkit has been designed to be extendable Users with some C C ex perience can write their own primitive blocks and compile them into the inbuilt library Chapter 6 explains the necessary steps Chapter 2 CTK Installation 2 1 Unix users 1 Download the gzip compressed CTK tar file This is available from http www dcs shef ac uk research groups spandh projects respite ctk 2 Unpack the archive if you have not done so already cd usr local gunzip CTKvi 1 1 tar gz uncompress the archive tar xf CTKvi 1 1 tar unpack it This creates the directory usr local CTK containing the files from the main archive If you have downloaded the data tar file then unpack this now aswell First copy it to usr local or whereever you wish to place CTK and then cd usr local gunzip CTK data gz uncompress the archive tar xf CTK data tar unpack it The rest of these instructions assume that CTK is installed in usr local CTK 3 Set some CTK environment variables in the file profile or login depending on your shell in your
7. these operations in one go instead we must build an intermediate block from the inbuilt blocks Delay Scale and Tee Figure 4 4 shows the architecture of this intermediate block So first we will construct the intermediate block shown in Figure 4 4 and give it a typename tap and then we will construct the delay line by arranging these tap blocks in series as in Figure 4 3 In any text editor open a new file and call it example2 ctk At the top of the file type usr local bin CTKScript Multiply and Delay Multiply and Delay Figure 4 3 DELAY SCALE KEY Parameter Input Socket Type Delay Type Scale D Output Socket Type tap Figure 4 4 Multiply and Delay 19 20 Next we need to start the description of a new block to which we will give the typename tap To do this we just type BLOCK tap Now we add the subblocks Delay Scale and Tee which go together to make up the tap block To do this we just add the lines ADD di Delay DELAY 20 ADD si Scale X 0 5 ADD t1 Tee These blocks need to be connected in series We do not need any explicit CONNECT statements because the default series connections will give the right result However once these three are connected there will still be a few loose connections namely the input of the Delay block and the two outputs of the Tee block These connections are intended to be inputs and outputs of the intermediate block itself This
8. CTKROUT scripts spectrogram ctk Now add the Spectrogram block at the end of the main block definition i e im mediately before the ENDBLOCK ADD s Spectrogram We do not need a CONNECT command as by default the Spectrogram block will take input from the proceeding Display block Display blocks are designed to both display their input and pass it on to their output Finally we want to display the output of the Spectrogram So we need to add another display block For Qt based output type ADD d2 Display Qt based output and for MATLAB output first add ADD d2 MDisplay BEFORE PLOT subplot 2 1 2 and then edit the definition of block d1 to be ADD di MDisplay BEFORE PLOT subplot 2 1 1 17 The BEF RE PLOT parameter of the MDisplay allows the MATLAB plot to be customised by issuing MATLAB commands that are executed immediately be fore the plot is generated In this case it is used to make the waveform and the spectrogram appear as subplots in a single window The script is now complete If you now go back to the Unix command shell and type examplei ctk 10 100 200 a window showing both the waveform and the spectrogram something like that shown in Figure 4 2 should appear If the script fails to run examine the error message and check your script against the script listed in Appendix A 1 di 3 T T T L L L L 0 100 200 300 400 500 600 700 800 900 1000 d2 400 2 300 5 200
9. FREQ 2 ADD i2 SineWave DURATION 1 SAMPLE RATE 1000 FREQ 3 ADD i3 SineWave DURATION 1 SAMPLE RATE 1000 FREQ 4 These SineWave blocks are generators as they have signal outputs 1 each but no signal inputs They have tunable parameters and in this case the parameters are set to produce sinusoids of 1 second duration at a sample rate of 1000 samples second and with frequencies of 2 3 and 4 Hz respectively We now need a way of superimposing these signals This is done using the inbuilt block called Adder This takes an arbitrary number of inputs and sums them to produce one output Type ADD a Adder NINPUTS 3 By default Adder expects two inputs In this case we have 3 sinusoids to add so the NINPUTS number of inputs parameter has been set to 3 We can now request a graphical display of the superimposed waveform The basic mechanism for graphical output is to use the inbuilt block Display which generates a Qt based output window An alternative for users with MATLAB is to use the MDisplay block which uses the CTK MATLAB interface to create a graphic in a MATLAB window So next type either ADD di Display Qt based output or ADD di MDisplay MATLAB based output As well as specifying the system s subblocks we must also describe how they are connected By default the toolkit will try to connect the blocks in series in the same order as they occur in the script If as in this case this is not what is wanted then connections m
10. Qt library If you re using GNU g you may also want to set these e LIBRARY PATH contains library file path e CPLUS INCLUDE PATH contains C include file path This is done like this In profile if your shell is bash ksh zsh or sh add the following lines QTDIR usr local qt PATH QTDIR bin PATH if LD_LIBRARY_PATH then LD_LIBRARY_PATH QTDIR 1ib LD_LIBRARY_PATH else LD LIBRARY PATH QTDIR lib fi LIBRARY PATH LD LIBRARY PATH if CPLUS INCLUDE PATH then CPLUS INCLUDE PATH QTDIR include CPLUS INCLUDE PATH else CPLUS INCLUDE PATH QTDIR include fi export QTDIR PATH MANPATH LD_LIBRARY_PATH LIBRARY PATH export CPLUS INCLUDE PATH In login in case your shell is csh or tesh add the following lines if QTDIR then setenv QTDIR usr local qt endif if 7PATH then setenv PATH QTDIR bin PATH else setenv PATH QTDIR bin endif if 7MANPATH then setenv MANPATH QTDIR man MANPATH else setenv MANPATH QTDIR man endif if LD LIBRARY PATH then setenv LD_LIBRARY_PATH QTDIR lib LD LIBRARY PATH else setenv LD_LIBRARY_PATH QTDIR lib endif if LIBRARY PATH then setenv LIBRARY PATH LD LIBRARY PATH endif if CPLUS INCLUDE PATH then setenv CPLUS INCLUDE PATH QTDIR include CPLUS INCLUDE PATH else setenv CPLUS INCLUDE PATH QTDIR include endif After you have done this you will need to login again or re source the profile bef
11. RESPITE CASA Toolkit CTK v1 1 1 User s Guide Jon Barker March 8 2001 Contents Introduction CTK Installation 2 l Unix USCS Re ee erde The CTK Design Writing CASA Toolkit Scripts 4 1 Tutorial 1 Displaying a spectrogram of superimposed sinusoids 4 2 Tutorial 2 Designing an intermediate level block 4 3 Tutorial 3 Missing Data Speech Recognition Using the CTK Graphical User Interface Extending the CTK Inbuilt Block Library 6 1 Setting Up Your Local 6 2 Writing Your Own Primitive Block LL Scripts for User Guide Tutorials A 1 Script for Tutorial llle A 2 Script for Tutorial2 A 3 Script for Tutorial 3 ow re 12 18 22 27 28 28 29 Chapter 1 Introduction The RESPITE project aims to extend and apply two novel technologies miss ing data theory and multi stream theory to the problem of robust automatic speech recognition ASR with particular application to cellular phones and in car environments Both these approaches rely on the successful identifica tion of regions of the signal spectrum containing reliable information Within the RESPITE project several contrasting approaches to this problem will be compared One of these approaches is Computational Auditory Scene Analysis The RESPITE CASA Toolkit aims to provide a flexible extensible and con sisten
12. TE 100 0 0 1 2 3 4 5 6 time Figure 4 2 CTK Data Flow This example also serves to illustrate an important feature of the data flow underlying the CASA toolkit Whereas the data entering the Spectrogramblock is a simple signal with time as the only dimension i e it is 1 dimensional the data leaving the Spectrogram block has two dimensions time and frequency All data must have time as a dimension but on top of that it may have any number of other dimensions Each inbuilt block is designed to examine the There is a corresponding AFTER_PLOT command that allows MATLAB commands to be executed immediately after the plot has been generated 18 dimensionality of the data entering it and to act appropriately So in this example the output of the first Display block is a simple graph with time along the x axis while the output of the second Display block is a colour image map with time and frequency along the x and y axes There is only one type of Display block but it behaviour alters according to the nature of its inputs 4 2 Tutorial 2 Designing an intermediate level block In the previous tutorial we saw how we can use an intermediate level block to perform the function of several connected inbuilt blocks We saw how an intermediate block called Spectrogram was used which was composed of the series combination of a Frame block and an FFT block Why go to the trouble of defining an intermediate blocks when we could just write it ou
13. cks process ing blocks and sink blocks A source block has output sockets but no input sockets A source block will generate its own data e g a pink noise generator block or will represent a file of stored data a WAV file reading block The data that is read or generated is passed to the source block s output socket A processing block has both input sockets and output sockets A processing block will take data from its inputs perform some operation on the data and pass the result to its output sockets For example an FFT block takes a frame START INC DURATION Type Output Type Ramp KEY Reset Compute Compute Parameter gt Input Socket SCALE b Output Socket Type Tee Type Multiplier Compute Compute Figure 3 1 A collection of primitive blocks Each primitive block embodies an algorithm Block s may also have input sockets output sockets and user parameters of data from its input performs an FFT and passes the transformed frame to its output A sink block is a block that has inputs but no output sockets Sink blocks generally represent output files or data displays System Sub components At the lowest level of description aCASA sytem is described by a set of primitive blocks their parameters and the connections that exist between them How ever when designing systems it may be more convenient to describe the system at a higher level as a set of sub components which ar
14. e themselves composed of primitive blocks This may be convenient if identical sub components are used more than once in the same system or if the user wishes to reuse identical sub components in other systems The toolkit allows the user to define a sub component or intermediate block in terms of the primitive blocks it is composed of Once defined this sub component may be used as though it were itself a primitive block This heirarchical sub component design may be extended so that higher level intermediate blocks may themselves be composed of mixtures of primitive blocks and lower level intermediate blocks Describing a system CASA systems and system sub components i e intermediate blocks may be described using a simple script like language and stored in files with the exten sion ctk A ctk file may contain the description of one or more sub components or a 10 DELAY SCALE KEY P Parameter Input Socket Type Delay Type Scale b Output Socket Figure 3 2 An intermediate level block defined in terms of primitive blocks complete system In either case the description includes a list of all the blocks contained in the system or sub component together with their parameter values and a description of the connections between their inputs and outputs Chapter 4 provides a tutorial describing in detail how to write CTK scripts Type Main KEY START INC DURATION T Parameter Type Ramp Type Output
15. ectory mkdir example3 cd example3 cp CTKROOT tutorial example3 ctk The script takes two arguments first the name of the ratemap file to be recog nised and second a string representing the correct transcription against which the result will be compared A set of ratemaps for TIDigits mixed with factory noise at various SNRs is distributed with the toolkit under the CTKROOT data directory So to test the script on the utterance 1159 at 10dB SNR type example3 ctk CTKROOT data factory 1159a 10 1159 After typing this there will be a short pause while the HMM files are read in Once the HMMs have been read a few lines appear summarising the HMM data After this recognition will commence and digits should appear on the screen as they are recognised When recognition is complete a line of statistics will appear summarising the systems performance The final output should look something like this 24 yelp 85 example3 ctk CTKROOT data factory 1159a 10 1159 num HMMs 12 num states 8 num mixes 10 vec size 64 num dist 960 1157 1159 Nin 4 Nout 4 H 3 I 0 D 0 S 1 Cor 75 Acc 75 Ni 4 Nout 4 H 3 I 0 D 0 S 1 Cor 75 Acc 75 This shows that the utterance 1159 was recognised as 1157 The statistics on the left show that 4 digits were read in and 4 digits were output there were 3 hits no insertions no deletions and 1 substitution Word correctness is 7596 and word accuracy is 75 The syste
16. he script simply type example2 ctk If all goes well two sinewaves should appear with one reduced in magnitude and delayed with respect to the other Using the MATLAB interface this should look something like Figure 4 5 If the script fails to run then you have probably made a typing mistake Examine the error message and check your script against the script listed in Appendix A 2 first output 0 5 T T T 0 4 0 5 1 1 1 1 1 1 1 0 50 100 150 200 250 300 350 400 450 500 second output 0 1 T T T T 0 1 i L L L L L L 0 50 100 150 200 250 300 350 400 450 500 Figure 4 5 4 3 Tutorial 3 Missing Data Speech Recogni tion In this final tutorial we will look at how the CASA toolkit can be used to run speech recognition experiments Figure 4 6 shows the block diagram for a system that performs missing data speech recognition At the bottom of the diagram is the block which performs 23 the missing data Viterbi decoding This is similar to a standard speech decoder but as well as taking a representation of the speech signal it also has an input called a mask which tells the decoder which elements of the representation may be considered to be reliable Typically some form of spectro temporal representation is used and the mask should indicate which spectro temporal elements have a favourable local SNR i e which elements are clean Following Figure 4 6 we see that the first opera
17. home directory Create the file if it is not there already e CTKROOT wherever you installed CTK e CTKLOCAL for user local CTK paths e PATH tolocate the CTK binaries e MANPATH to access the CTK man pages This is done like this In profile if your shell is bash ksh zsh or sh add the following lines CTKROOT usr local CTK CTKLOCAL usr local CTK local PATH CTKLOCAL bin PATH if MANPATH then MANPATH CTKROOT man MANPATH else MANPATH CTKROOT man fi export CTKROOT CTKLOCAL PATH MANPATH In login in case your shell is csh or tcsh add the following lines if CTKROOT then setenv CTKROOT usr local CTK endif if CTKLOCAL then setenv CTKLOCAL CTKROOT local endif if PATH then setenv PATH CTKLOCAL bin PATH else setenv PATH CTKLOCAL bin endif if 7MANPATH then setenv MANPATH CTKROOT man MANPATH else setenv MANPATH CTKROOT man endif After you have done this you will need to login again or re source the profile before continuing so that at least CTKROOT and CTKLOCAL are set The installation will give an error message and not proceed otherwise Make sure Qt is properly installed The Qt widget library is needed if you want to run the CTK GUI It should be installed on your system and the following environment variables should be set e QTDIR wherever Qt is installed on your system e PATH to locate the moc program e LD_LIBRARY_PATH for the shared
18. m seems to work well with this single utterance but to really test it we need to run it over a large test set of several hundred utterances One approach would be to write a shell script to repeatedly execute the example3 ctk com mand each time with a different ratemap file and a different transcription string However there are two problems with this First reading in the HMMs repre sents a sizeable computational overhead We do not want to have to reread the HMMs for every utterance to be recognised Second there is no convenient way of calculating the overall performance if we have invoked example3 ctk sepa rately for each utterance to be recognised In order to overcome these problems we can use the CTK script argument list mechanism Script argument lists A script argument list is a text file where each line supplies the arguments for a separate invokation of the block process described by the script These argument lists are introduced to the script on the command line using CTKScript s S option If a script requires more than one argument these arguments can all be specified in a single argument list containing several columns or they can be split across multiple argument lists with fewer columns When more than one argument list is used each list is introduced on the command line with a separate S An example makes this clear Make your window as wide as possible and then try the following command example3 ctk S CTKROOT data flis
19. main block will have a SineWave source block then a series of tap blocks Add to following lines BLOCK main ADD ii SineWave DURATION 0 5 FREQ 5 SAMPLE RATE 1000 ADD tapi tap ADD tap2 tap ADD tap3 tap ADD tap4 tap ADD tap5 tap The default connections will string these blocks together in series Now lets display the signal at a couple of different points along the delay line To do this we simply add a couple of display blocks and connect them to the delay line using the spare outputs of the tap blocks For example if you are using the MATLAB interface add the following ADD first output MDisplay BEF RE PLOT subplot 2 1 1 ADD second output MDisplay BEFORE_PLOT subplot 2 1 2 or if you are using the Qt based graphical output add ADD first output Display ADD second output Display If you have no graphical output you can monitor the output on the terminal using ADD first output utput ADD second output utput Then to connect the outputs to the delay line add the following CONNECT tapi out2 first output CONNECT tap4 out2 second output Here we have connected the display to tapi and tap4 but they could equally well be connected at any two other positions And finally finish the main block definition by adding ENDBLOCK 22 The example is now ready to be run To do this you first need to make sure the script file is executable In the Unix command shell type chmod u x example2 ctk Then to run t
20. must be made explicit and is done so by use of the INPUT and OUTPUT commands To handle the input add the line INPUT ini d1 This declares that the intermediate has an input socket and this socket feeds into the input of the subblock called d1 i e it feeds into the Delay block Likewise for the outputs we need to add the lines OUTPUT outi ti outi OUTPUT out2 t1 out2 This declares that the intermediate block has two output sockets and they are respectively output socket 1 and output socket 2 of the block called t1 i e the Tee block We can now complete the block definition with the line ENDBLOCK With this definition the Tap block has no parameters The delay and scale factor will be fixed at 20 and 0 5 as stated in the definition of its Delay and Scale subblocks If we want tap blocks with adjustable delays and scale factors we must make the parameters of its subblocks visible as parameters of the tap block itself see on Figure 4 4 how tap has a parameter DELAY that is linked to the parameter DELAY of its subblock Delay To do this we use the PARAMETER command So just before the ENDBLOCK insert the lines PARAMETER DELAY d1 DELAY PARAMETER SCALE s1 X 21 This gives the tap block parameters DELAY and SCALE and attaches these to the DELAY parameter of the Delay block and the X parameter of the Scale block respectively Now that the tap block has been defined it is ready to be used in the construction of the main block The
21. n your system but it should point to the location of the files libmat so and libeng so 7 Preparing the CTK script files The example CTK script files need to be editted so that the first line of each script references the CTKScript script interpreter binary This can be done automatically by using the fix_scripts script Simply type fix scripts 8 Setting up your editor emacs only When viewing CTK script files emacs should be put into sh mode To make this happen automatically add the following lines to your emacs file setq auto mode alist append ctk sh mode auto mode alist Chapter 3 The CTK Design Bottom up CASA systems are typically built in a modular fashion from fairly standard components The CASA Toolkit aims to ease the design of CASA systems by offering the user a library of well documented pre defined processing blocks and a mechanism for combining them in an arbitrary fashion Primitive Blocks At the heart of the CASA toolkit is a library of primitive blocks These blocks are defined in terms of the foll owing e Input sockets e Output sockets e An algorithm describing how data at the inputs is to be transformed and placed on the outputs e Parameters A set of parameters which may be set from outside the block and may control the block s operation Figure 3 1 illustrates several primitive blocks Block s may be broadly divided into 3 distinct types source blo
22. ore continuing so that at least QTDIR is set The installation will give an error message and not proceed otherwise Compile the CTK binaries You may find that pre compiled binaries are available for download di rectly from the CTK web site In this case simply copy the tar file to CTKROOT and unpack it If you can t find binaries for your system then you will have to compile your own The CTKROOT src directory contains a makefile which compiles the CTK binaries The basic system with no MATLAB or Qt support can be built by typing cd CTKROOT src make If this fails or if you want to install the MATLAB interface or the Qt based GUI you will have to edit CTKROOT src Makefile First read the INSTALL file found in CTKROOT src Setting up with MATLAB support If you have compiled with MATLAB support you will also need to set your LD_LIBRARY_PATH to locate the MATLAB libraries This is done like this In profile if your shell is bash ksh zsh or sh add the following lines if LD LIBRARY PATH then LD LIBRARY PATH usr local matlab extern lib sol2 LD LIBRARY PATH else LD LIBRARY PATH usr local matlab extern lib sol12 In login in case your shell is csh or tcsh add the following lines if LD LIBRARY PATH then setenv LD_LIBRARY_PATH usr local matlab extern lib sol2 LD LIBRARY PATH else setenv LD_LIBRARY_PATH usr local matlab extern lib sol2 endif The exact path may vary o
23. rent utterance and those on the right are a running total showing the performance so far The final recognition accuracy after all 240 utterances have been processed should be about 8496 26 AUInputFile Floor X 0 0 Power X 2 0 Scale X 2 0 amp HMMs q Transcriptions Recognition Statistics Power X 2 0 HMMDecoderMD Figure 4 6 Chapter 5 Using the CTK Graphical User Interface MORE STUFF HERE Chapter 6 Extending the CTK Inbuilt Block Library This chapter provides some brief notes to help user s who wish to add their own primitive blocks to the toolkit 6 1 Setting Up Your Local CTK Before you add your own blocks to CTK you have to make your own local version of the CTK binaries This is done by following the steps 1 First you have to make your own local CTK directory For example mkdir home jon CTK 2 Then if your shell is csh or tcsh edit your login and change the definition of CTKLOCAL to point to this directory setenv CTKLOCAL home jon CTK If your shell is bash ksh zsh or sh then edit your profile to include the line CTKLOCAL usr local CTK local 3 Now copy the files under CTKROOT local to CTKLOCAL cp r CTKROOT local CTKLOCAL 4 Now test that it is working by trying to rebuild the binaries locally 28 29 cd CTKROOT src make 5 Finally edit your PATH to add the CTKLOCAL bin to your path making sure it precedes CTKROOT
24. t framework for the development and testing of CASA systems within the project This framework should facilitate the smooth integration of CASA software components contributed by the various RESPITE partners It is also hoped that this toolkit may prove a useful resource to the CASA community as a whole both as a reference framework for comparing different developments and as a way to lower the barriers to entry into this research field To allow the desired flexibility and extensibility the CASA Toolkit has been designed around a block processing paradigm Much of this design has been influenced by approaches taken in other signal processing systems notably the Speech Training and Recognition Unified Tool STRUT of the TCTS lab at Facult Polytechnique de Mons and the Ptolemy project Department of EECS UC Berkeley The structure of the rest of the manual Chapter 2 explains how to download and install the CTK package on your system Chapter 3 describes the block orientated processing and online data flow model underlying the toolkit Chapter 4 provides a number of walk through tutorials which illustrate how to write CASA Toolkit scripts to run simple signal processing systems These tutorials have been designed to be progressive and to illustrate as many of the principles of the CTK package as possible Once you have worked through the tutorials you should be in a position to start building your own CTK systems Chapter 5 describes
25. t longhand in terms of inbuilt blocks The obvious advantage is that it simplifies scripts making them both shorter and easier to read and write However a far more important advantage of intermediate blocks is that they allow a degree of design reuse Intermediate blocks definitions can be written and stored in a library These intermediate blocks can then be used as parts of a more complex system through use of the INCLUDE mechanism So a simpler unit such as a spectrogram can be designed once and then reused in any number of more complex systems It should be noted that an intermediate level block may itself be designed in terms of other intermediate level blocks and there is no limit to the level of such nesting that may be used In the previous tutorial we used the pre defined Spectrogramintermediate block that is provided as part of th CTK intermediate block library All we had to do was add the appropriate INCLUDE line at the top of our script file then we could use Spectrogram exactly as if it was one of the toolkit inbuilt blocks In this tutorial we will see how to define our own intermediate level blocks so that they can be used like inbuilt blocks in this way In this tutorial we will construct a lossy delay line like that shown in Figure 4 3 Each block in the delay line performs a delay operation a scaling operation and a teeing operation to split the data flow into two outputs There is no one inbuilt block that can perform all
26. t which superimposes three sinusoidal signals and displays the spectrogram of the resulting signal In any text editor open a new file and call it examplel ctk At the top of the file type usr local bin CTKScript This line tells the shell that the file is a script and should be interpreted by the CTKScript command in usr local bin If your CTK binaries are installed somewhere other than usr local bin then make sure you use the appropriate path Next start the description of a new processing block On a new line type BLOCK main This creates a new block and assigns it the name main A script file can define any number of blocks The CTKScript command takes as an optional argument the name of the block to execute By default it will search for a block called 12 13 main If the script contains only one top level block then it makes sense to call this block main so that CTKScript will work using the default block name argument The top level block main describes the entire system and is composed of a number of interconnected lower level blocks These lower level blocks can be either intermediate level blocks themselves composed of a network of lower level blocks or one of the fixed library of lowest level inbuilt blocks For this example we must start out by adding three inbuilt sine wave generator blocks To do this we use the CTK script ADD command Type ADD ii SineWave DURATION 1 SAMPLE RATE 1000
27. tion is to read the noisy speech from an AU sound file In this example recognition will be based on a ratemap representation The ratemap is a kind of auditory inspired spectrogram and is formed by passing the signal through a gammatone filterbank and then perform ing some leaky integration and downsampling The ratemap is then duplicated using a tee block and one copy passed directly to the decoder and the other copy is passed down the right hand side of the figure through the blocks that generate the mask The mask is generated by performing a simple noise estimation and then using a comparator to set the mask to true at the spectro temporal points where the estimated clean signal i e the noisy signal after subtraction of the estimated noise makes up over half the energy of the noisy signal The mask is then passed into the second input of the decoder The CTK script to describe this process is listed in Appendix A 2 The script shown differs slightly from the diagram in that it omits the initial ratemap generation stage and instead reads in precomputed ratemaps Note that the default series block connection means that very few of the block connections need to be specified explicitly We will now see how this script can be used both for recognising a single ut terance or recognising a small corpus of test utterances First make a new directory called example3 and copy the example3 CTK script supplied with the distribution into this dir
28. ts test240 10 flist S CTKROOT data transcripts transcripts 240 The first S parameter introduces the argument list containing the paths of the complete test set of 240 files to be recognised The second argument list is a list of the correct transcriptions of these 240 utterances Note all the argument lists should have the same length if they do not then an error will be reported This command should produce output that starts like that below ye1p 333 example3 ctk S CTKROOT data flists test240 10 flist S CTKROOT data transcripts transcripts 240 num HMMs 12 num states 8 num mixes 10 vec size 64 num dist 960 1157 1159 Nin 4 Nout 4 H 3 I 0 D 0 S 1 Cor 75 Acc 75 Nin 4 Nout 4 H 3 I 0 0 S 1 Cor 75 Acc 75 12773073 1273073 Nin 7 Nout 8 H 7 I 1 D 0 S 0 Cor 100 Acc 85 7143 Nin 11 Nout 12 H 10 I 1 D 0 S 1 Cor 90 9091 Acc 81 8182 1627 127 Nin 3 Nout 4 H 3 I 1 D 0 S 0 Cor 100 Acc 66 6667 Nin 14 Nout 16 H 13 I 2 D 0 S 1 Cor 92 8571 Acc 78 5714 61280 1280 Nin 4 Nout 5 H 4 I 1 D 0 S 0 Cor 100 Acc 75 Nin 18 Nout 21 H 17 I 3 D 0 S 1 Cor 94 4444 Acc 77 7778 12 12 Nin 2 Nout 2 H 2 I 0 D 0 S 0 Cor 100 Acc 100 Nin 20 Nout 23 H 19 I 3 D 0 S 1 Cor 95 Acc 80 25 To recognise the full set of 240 utterances may take some time The recognition can be halted at any time by using the Ct1 C key The statistics on the left of the screen apply to the cur
29. ust be specified explicitly In this example each of the sine wave generators must connect to one of the three inputs on the adder block and the output of the added block must connect to the display block To make these connections add the lines CONNECT ii outi a ini 14 CONNECT i2 out1 a in2 CONNECT i3 out1 a in3 The first line connects the output socket named out1 of the block called 11 to the input socket named in1 of the Adder block called a The second and third lines likewise connect inputs i2 and 13 to the 2nd and 3rd inputs of the Adder block a Note for all blocks the input sockets will be named in1 in2 etc and the output sockets named out1 out2 etc If a block has only 1 input socket or one output socket then the socket name can be omitted from the CONNECT statement without any ambiguity So in this case we could more simply write CONNECT 11 a int CONNECT i2 a in2 CONNECT i3 a in3 To connect the adder block to the display block we could add the line CONNECT a di But in this case there is no need to make this connection explicitly as it follows the default behaviour of connecting blocks in series in the order in which they are defined in the script So this line can be omitted Finally add the line ENDBLOCK This marks the end of a block definition We have now defined a simple system that will add three sinusoids and display the resulting waveform This system is illustrated in Figure 4 1 We can now exec
30. ute the script and see the result From the editor first save the script and then in the Unix shell window make the script executable by using the Unix chmod command chmod u x examplei ctk Now simply execute it by typing examplei ctk TAt present in the case of multiple input output sockets it may be necessary to refer to the block s documentation to see how to correctly connect up the block In future versions nicknames may be associated with these raw input and output names to make their meaning more transparent 15 Duration Samplerate Fre ES Name main Type SineWave Name il Type SineWave Type Adder Type Display Name i2 Name a Name dl Type SineWave Name i3 Figure 4 1 A window should appear displaying a periodic waveform If the MATLAB interface display is used i e MDisplay there will be a short pause while the MATLAB engine starts up The waveform is the addition of sinusoids with frequencies of 2 3 and 4 Hz By changing the FREQ parameters of the SineWave blocks and rerunning the script you can see the effect of adding sinusoids with different frequencies Alterna tively you can set the sinusoid frequencies from the command line To do this you must first edit the script so the SineWave blocks are defined as ADD ii SineWave DURATION 1 SAMPLE RATE 1000 FREQ 1 ADD i2 SineWave DURATION 1 SAMPLE RATE 1000 FREQ 2 ADD i3 SineWave DURATION 1 SAMPLE RATE 1000 FREQ 3 The 1 2 and
Download Pdf Manuals
Related Search