SpkAdaptMain.cpp

Go to the documentation of this file.

/*
This file is part of LIA_RAL which is a set of software based on ALIZE
toolkit for speaker recognition. ALIZE toolkit is required to use LIA_RAL.

LIA_RAL project is a development project was initiated by the computer
science laboratory of Avignon / France (Laboratoire Informatique d'Avignon -
LIA) [http://lia.univ-avignon.fr <http://lia.univ-avignon.fr/>]. Then it
was supported by two national projects of the French Research Ministry:
        - TECHNOLANGUE program [http://www.technolangue.net]
        - MISTRAL program [http://mistral.univ-avignon.fr]

LIA_RAL is free software: you can redistribute it and/or modify
it under the terms of the GNU Lesser General Public License as
published by the Free Software Foundation, either version 3 of
the License, or any later version.

LIA_RAL is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public
License along with LIA_RAL.
If not, see [http://www.gnu.org/licenses/].

The LIA team as well as the LIA_RAL project team wants to highlight the
limits of voice authentication in a forensic context.
The "Person Authentification by Voice: A Need of Caution" paper
proposes a good overview of this point (cf. "Person
Authentification by Voice: A Need of Caution", Bonastre J.F.,
Bimbot F., Boe L.J., Campbell J.P., Douglas D.A., Magrin-
chagnolleau I., Eurospeech 2003, Genova].
The conclusion of the paper of the paper is proposed bellow:
[Currently, it is not possible to completely determine whether the
similarity between two recordings is due to the speaker or to other
factors, especially when: (a) the speaker does not cooperate, (b) there
is no control over recording equipment, (c) recording conditions are not
known, (d) one does not know whether the voice was disguised and, to a
lesser extent, (e) the linguistic content of the message is not
controlled. Caution and judgment must be exercised when applying speaker
recognition techniques, whether human or automatic, to account for these
uncontrolled factors. Under more constrained or calibrated situations,
or as an aid for investigative purposes, judicious application of these
techniques may be suitable, provided they are not considered as infallible.
At the present time, there is no scientific process that enables one to
uniquely characterize a persones voice or to identify with absolute
certainty an individual from his or her voice.]

Copyright (C) 2004-2010
Laboratoire d'informatique d'Avignon [http://lia.univ-avignon.fr]
LIA_RAL admin [alize@univ-avignon.fr]
Jean-Francois Bonastre [jean-francois.bonastre@univ-avignon.fr]
*/

#include <iostream>
#include "SpkAdapt.h"
#define MANDATORY true
#define OPTIONAL false
#define ARG_REQUIRED true
#include "liatools.h"

int main(int argc, char *argv[])
{
  try
  {
    ConfigChecker cc;
    cc.addIntegerParam("verboseLevel", false, true,
      "level of the verbose information 0=no verbose, 1=normal, 2=more");
    cc.addStringParam("config", OPTIONAL, ARG_REQUIRED,
      "Train config file name ");
    cc.addStringParam("configTest", OPTIONAL, ARG_REQUIRED,
      "Unsupervised adaptation config file name ");
    cc.addStringParam("targetIdList", MANDATORY, ARG_REQUIRED,
      "targets list");
    cc.addStringParam("inputWorldFilename", MANDATORY, ARG_REQUIRED,
      "world model");
    cc.addStringParam("testsIdList", MANDATORY, ARG_REQUIRED, "tests list");
    cc.addBooleanParam("NISTprotocol", MANDATORY, ARG_REQUIRED,
      "true to follow unsupervised NIST protocol, false to follow BATCH protocol");
    cc.addBooleanParam("WMAP", false, false,
      "NO MORE USED, Choice of WMAP (one Gaussian) to compute trial feature server weights, need other param : MUtarget,MUimp,SIGMAtarget and SIGMAIMP;TAR weight and IMPweight, thrMin and thrMax");
    cc.addBooleanParam("WMAPGMM", false, false,
      "Choice of WMAP GMM to compute trial feature server weights, param needed : two gmm of scores, TARScoresModel and NONScoresModel. Need also a value for Priors update, initPriorImp (10) and initPriorTar(1) ");
    cc.addBooleanParam("TNORM", false, false,
      " TNORM scores before computing WMAP, need gmm of Tnormed scores and a param : impScoreFile, score file for TNORM and testsNames, the list of trial segments ");
    cc.addBooleanParam("FAST", false, false,
      "FAST MODE : For computing LLR use top ten info files AND FOR FUSING EM MODELS WHEN UPDATING");
    cc.addBooleanParam("FromResFile", false, false,
      " Avoid to compute LLR for WMAP computing, search LLR in a score file, param needed: InputResFilename");
    cc.addBooleanParam("CrossValid", false, false,
      " compute LLR of a percentage of train data on a model learnt on a percentage of train data AverageIt times and return the best ML (one iter) model, need AverageIt, SelectedTrain.");
    cc.addBooleanParam("Oracle", false, false,
      " For Oracle (supervised) experiments, need targetTests to perform adaptation on target tests only");
    cc.addBooleanParam("REGRESS", false, false,
      "Use logistic regression for adaptation weights computing, need BETA and THETA");
    cc.addIntegerParam("MaxMixturesCount", false, true,
      "max mixtures stored in the mixture server (in memory)");
    cc.addIntegerParam("LLKthreshold", false, true,
      "thresholds the results of LLR in order to avoid problem in WMAP computing");
    cc.addStringParam("computeLLKWithTopDistribs", true, true,
      "PARTIAL/COMPLETE: will compute LLK with topdistribs. COMPLETE: add world LLK to client LLK, PARTIAL: just keeps the topDistrib LLK");
    cc.addIntegerParam("topDistribsCount ", false, true,
      "Number of distrib to approximate complete LLK");
    cc.addStringParam("InfoExtension", false, true,
      "Extension for top ten info files");
    cc.addStringParam("InfoPath", false, true, "Path for top ten info files");
    cc.addFloatParam("OptimalScore", false, true,
      "Used for updating  priors (fixedPriors == false), if LLR is superior add one for target count, else add one for impostors count ");
    cc.addBooleanParam("SegMode", MANDATORY, ARG_REQUIRED,
      "For adaptation by segments (for the moment support only 3 frame segments)");
    cc.addIntegerParam("mixtureDistribCountForGMMScores", false, true,
      "NUMBER OF GAUSSIAN IN SCORES GMM");
      cc.addBooleanParam("ScoresByTarget", MANDATORY, ARG_REQUIRED," IF SET TO TRUE TWO GMM (TAR AND NON) ARE NEEDED BY CLIENT (idcinet.tar and idclient.non)");
    cc.addStringParam("TARListFilename", false, true, "List of the TAR model id (used when ScoresByTarget is set to true)");
cc.addIntegerParam("ResetNbAdapt ", false, true,
      "Max number of tests used for adaptation");
    Config tmp;
    CmdLine cmdLine(argc, argv);
    if (cmdLine.displayHelpRequired())
      {                         // --help
        cout << "************************************" << endl;
        cout << "********** SpkAdapt.exe *********" << endl;
        cout << "************************************" << endl;
        cout << endl;
        cout <<
          "Unsupervised adaptation process, update model using test trial information following NIST protocol"
          << endl;
        cout << "" << endl;
        cout << endl;
        cout << "Command Line example: " << endl;
        cout <<
          "SpkAdapt.exe --config cfg/TrainTarget.cfg --configTest cfg/TrainTest.cfg --inputWorldFilename world --targetIdList ./lst/id.lst --testsIdList ./lst/tests.list --outputLLRFilename file --NISTprotocol true"
          << endl;
        cout << cc.getParamList() << endl;
      }
    else
      {
        cmdLine.copyIntoConfig(tmp);
        Config config(tmp.getParam("config"));
        cmdLine.copyIntoConfig(config);
        Config configTest(tmp.getParam("configTest"));
        if (config.existsParam("verbose"))
          verbose = config.getParam("verbose").toBool();
        else
          verbose = false;
        bool segmode=false;
        if (config.existsParam("SegMode"))
          segmode = config.getParam("SegMode").toBool();
        if (verbose)
          verboseLevel = 1;
        else
          verboseLevel = 0;
        if (config.existsParam("verboseLevel"))
          verboseLevel = config.getParam("verboseLevel").toLong();
        if (verboseLevel > 0)
          verbose = true;
        
        TrainTargetAdapt(config, configTest);   // Launch process

      }
  }
  catch(alize::Exception & e)
  {
    cout << e.toString() << endl;
  }

  return 0;
}