00001 /*!@file Neuro/InferoTemporalHmax.H Object recognition module */ 00002 00003 // //////////////////////////////////////////////////////////////////// // 00004 // The iLab Neuromorphic Vision C++ Toolkit - Copyright (C) 2001 by the // 00005 // University of Southern California (USC) and the iLab at USC. // 00006 // See http://iLab.usc.edu for information about this project. // 00007 // //////////////////////////////////////////////////////////////////// // 00008 // Major portions of the iLab Neuromorphic Vision Toolkit are protected // 00009 // under the U.S. patent ``Computation of Intrinsic Perceptual Saliency // 00010 // in Visual Environments, and Applications'' by Christof Koch and // 00011 // Laurent Itti, California Institute of Technology, 2001 (patent // 00012 // pending; application number 09/912,225 filed July 23, 2001; see // 00013 // http://pair.uspto.gov/cgi-bin/final/home.pl for current status). // 00014 // //////////////////////////////////////////////////////////////////// // 00015 // This file is part of the iLab Neuromorphic Vision C++ Toolkit. // 00016 // // 00017 // The iLab Neuromorphic Vision C++ Toolkit is free software; you can // 00018 // redistribute it and/or modify it under the terms of the GNU General // 00019 // Public License as published by the Free Software Foundation; either // 00020 // version 2 of the License, or (at your option) any later version. // 00021 // // 00022 // The iLab Neuromorphic Vision C++ Toolkit is distributed in the hope // 00023 // that it will be useful, but WITHOUT ANY WARRANTY; without even the // 00024 // implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR // 00025 // PURPOSE. See the GNU General Public License for more details. // 00026 // // 00027 // You should have received a copy of the GNU General Public License // 00028 // along with the iLab Neuromorphic Vision C++ Toolkit; if not, write // 00029 // to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, // 00030 // Boston, MA 02111-1307 USA. // 00031 // //////////////////////////////////////////////////////////////////// // 00032 // 00033 // Primary maintainer for this file:Sophie Marat 00034 // $HeadURL: svn://isvn.usc.edu/software/invt/trunk/saliency/src/Neuro/InferoTemporalSIFT.H $ 00035 // $Id: InferoTemporalHmax.H 14244 2010-11-17 02:56:14Z sophie $ 00036 // 00037 00038 #ifndef INFEROTEMPORALSIFT_H_DEFINED 00039 #define INFEROTEMPORALSIFT_H_DEFINED 00040 00041 00042 #include "Component/ModelComponent.H" 00043 #include "Component/ModelParam.H" 00044 #include "Neuro/InferoTemporal.H" 00045 #include "SIFT/Keypoint.H" 00046 #include "SIFT/VisualObject.H" 00047 #include "SIFT/VisualObjectDB.H" 00048 #include "Channels/InputFrame.H" 00049 #include "Image/Pixels.H" 00050 #include "Image/Image.H" 00051 #include "Media/TestImages.H" 00052 00053 #include "Channels/RawVisualCortex.H" 00054 #include"Component/RawGistEstimatorStd.H" 00055 00056 #include "Neuro/SVMClassifierModule.H" 00057 00058 class Brain; 00059 00060 namespace nub { template <class T> class ref; } 00061 00062 00063 // ###################################################################### 00064 //! Inferotemporal interface using SIFT with feature learning 00065 /*! Model of the Inferior Temporal Cortex utilizing SIFT object recognition */ 00066 class InferoTemporalSIFT : public InferoTemporal 00067 { 00068 00069 public: 00070 //! Constructor 00071 InferoTemporalSIFT(OptionManager& mgr, 00072 const std::string& descrName = "Infero Temporal SIFT", 00073 const std:: string& tagName = "InferoTemporalSIFT"); 00074 00075 00076 00077 //! Destructor 00078 virtual ~InferoTemporalSIFT(); 00079 00080 std::string getObjNameAtLoc(const std::vector<TestImages::ObjData> &objects, const Point2D<int>& loc); 00081 00082 00083 00084 protected: 00085 //! Object recognition 00086 virtual void attentionShift(SimEventQueue& q, const Point2D<int>& location); 00087 00088 protected: 00089 OModelParam<std::string> itsSIFTStoredDatabase; //!< dir of stored SIFT Database 00090 OModelParam<std::string> itsITCMode; //!< option for training or testing mode 00091 OModelParam<std::string> itsPathMatch; //!< Path of the objects to consider for matching 00092 OModelParam<bool> itsCoarseReco; //!< Use the gist to do a coares pre-recognition 00093 00094 OModelParam<std::string> itsTrainSVM; //!< file where the training exemple for the SVM will be saved 00095 OModelParam<std::string> itsSVMId; //!< The Id of the object to train the SVM 00096 OModelParam<std::string> itsSVMClass; //!< The Class of the object to train the SVM - Used for completing the ID:Class table 00097 OModelParam<std::string> itsSVMModel; //!< file where the training exemple for the SVM will be saved The SVM model to use 00098 OModelParam<std::string> itsSVMRange; //!< file where the training exemple for the SVM will be saved The range to rescale data before SVM 00099 OModelParam<std::string> itsNameObj; //!< Name on the object that is processed to train SVM /// 00100 OModelParam<std::string> itsRecoSave; //!< file where the result of the recognition will be saved 00101 OModelParam<std::string> itsTable; //!< file where the table of the class and their Id is saved 00102 00103 00104 rutz::shared_ptr<VisualObjectDB> itsObjectDB; //!< our object database 00105 rutz::shared_ptr<VisualObjectDB> itsnewObjectDB; //!< to adapt the database to select only object to match within the given path 00106 rutz::shared_ptr<std::map<double,int> > itsPDFGist; //!< proba density of the recognition using gist 00107 00108 //GIST// 00109 nub::ref<RawVisualCortex> itsVisualCortex; 00110 //rutz::shared_ptr<RawVisualCortex> itsVisualCortex; // raw visual cortex use to compute coarse recognition with gist 00111 nub::ref<RawGistEstimatorStd> itsGistEstim; 00112 00113 00114 //nub::ref<SVMClassifierModule> itsClassifier; 00115 00116 //! (re-)initialization - load object database 00117 virtual void start1(); 00118 00119 //! destruction - save object database 00120 virtual void stop1(); 00121 00122 void getObjDBToMatch(const char *dir); 00123 00124 void computeGist(Image<PixRGB<byte> >); 00125 //Image<float> computeGist(Image<PixRGB<byte> >); 00126 00127 void gistSelect(std::map<double, int>::reverse_iterator iteratorPDFGist); 00128 00129 }; 00130 00131 #endif 00132 00133 00134 // ###################################################################### 00135 /* So things look consistent in everyone's emacs... */ 00136 /* Local Variables: */ 00137 /* indent-tabs-mode: nil */ 00138 /* End: */ 00139