Surprise.C

Go to the documentation of this file.

// ///////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
// JeVois Smart Embedded Machine Vision Toolkit - Copyright (C) 2016 by Laurent Itti, the University of Southern
// California (USC), and iLab at USC. See http://iLab.usc.edu and http://jevois.org for information about this project.
//
// This file is part of the JeVois Smart Embedded Machine Vision Toolkit.  This program is free software; you can
// redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software
// Foundation, version 2.  This program 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 General Public
// License for more details.  You should have received a copy of the GNU General Public License along with this program;
// if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
//
// Contact information: Laurent Itti - 3641 Watt Way, HNB-07A - Los Angeles, CA 90089-2520 - USA.
// Tel: +1 213 740 3527 - itti@pollux.usc.edu - http://iLab.usc.edu - http://jevois.org
// ///////////////////////////////////////////////////////////////////////////////////////////////////////////////////
/*! \file */
 
#include <jevoisbase/Components/Saliency/Surprise.H>
 
// ##############################################################################################################
Surprise::Surprise(std::string const & instance) :
    jevois::Component(instance)
{
  itsSaliency = addSubComponent<Saliency>("saliency");
}
 
// ##############################################################################################################
Surprise::~Surprise()
{ }
 
 
// ##############################################################################################################
namespace
{
// ######################################################################
// ##### Fast Log Function
// ######################################################################
  static const double TBL_LOG_P[] = {
    /* ONE   */  1.0,
    /* TWO52 */  4503599627370496.0,
    /* LN2HI */  6.93147180369123816490e-01,        /* 3fe62e42, fee00000 */
    /* LN2LO */  1.90821492927058770002e-10,        /* 3dea39ef, 35793c76 */
    /* A1    */ -6.88821452420390473170286327331268694251775741577e-0002,
    /* A2    */  1.97493380704769294631262255279580131173133850098e+0000,
    /* A3    */  2.24963218866067560242072431719861924648284912109e+0000,
    /* A4    */ -9.02975906958474405783476868236903101205825805664e-0001,
    /* A5    */ -1.47391630715542865104339398385491222143173217773e+0000,
    /* A6    */  1.86846544648220058704168877738993614912033081055e+0000,
    /* A7    */  1.82277370459347465292410106485476717352867126465e+0000,
    /* A8    */  1.25295479915214102994980294170090928673744201660e+0000,
    /* A9    */  1.96709676945198275177517643896862864494323730469e+0000,
    /* A10   */ -4.00127989749189894030934055990655906498432159424e-0001,
    /* A11   */  3.01675528558798333733648178167641162872314453125e+0000,
    /* A12   */ -9.52325445049240770778453679668018594384193420410e-0001,
    /* B1    */ -1.25041641589283658575482149899471551179885864258e-0001,
    /* B2    */  1.87161713283355151891381127914642725337613123482e+0000,
    /* B3    */ -1.89082956295731507978530316904652863740921020508e+0000,
    /* B4    */ -2.50562891673640253387134180229622870683670043945e+0000,
    /* B5    */  1.64822828085258366037635369139024987816810607910e+0000,
    /* B6    */ -1.24409107065868340669112512841820716857910156250e+0000,
    /* B7    */  1.70534231658220414296067701798165217041969299316e+0000,
    /* B8    */  1.99196833784655646937267192697618156671524047852e+0000,
  };
 
#define        ONE   TBL_LOG_P[0]
#define        TWO52 TBL_LOG_P[1]
#define        LN2HI TBL_LOG_P[2]
#define        LN2LO TBL_LOG_P[3]
#define        A1    TBL_LOG_P[4]
#define        A2    TBL_LOG_P[5]
#define        A3    TBL_LOG_P[6]
#define        A4    TBL_LOG_P[7]
#define        A5    TBL_LOG_P[8]
#define        A6    TBL_LOG_P[9]
#define        A7    TBL_LOG_P[10]
#define        A8    TBL_LOG_P[11]
#define        A9    TBL_LOG_P[12]
#define        A10   TBL_LOG_P[13]
#define        A11   TBL_LOG_P[14]
#define        A12   TBL_LOG_P[15]
#define        B1    TBL_LOG_P[16]
#define        B2    TBL_LOG_P[17]
#define        B3    TBL_LOG_P[18]
#define        B4    TBL_LOG_P[19]
#define        B5    TBL_LOG_P[20]
#define        B6    TBL_LOG_P[21]
#define        B7    TBL_LOG_P[22]
#define        B8    TBL_LOG_P[23]
 
  static const double _TBL_log[] = {
    9.47265623608246343e-02, 1.05567010464380857e+01, -2.35676082856530300e+00,
    9.66796869131412717e-02, 1.03434344062203838e+01, -2.33635196153499791e+00,
    9.86328118117651004e-02, 1.01386139321308306e+01, -2.31635129573594156e+00,
    1.00585936733578435e-01, 9.94174764856737347e+00, -2.29674282498938709e+00,
    1.02539062499949152e-01, 9.75238095238578850e+00, -2.27751145544242561e+00,
    1.04492186859904843e-01, 9.57009351656812157e+00, -2.25864297726331742e+00,
    1.06445312294918631e-01, 9.39449543094380957e+00, -2.24012392529694537e+00,
    1.08398437050250693e-01, 9.22522526350104144e+00, -2.22194160843615762e+00,
    1.10351562442130582e-01, 9.06194690740703912e+00, -2.20408398741152212e+00,
    1.12304686894746625e-01, 8.90434787407592943e+00, -2.18653968262558962e+00,
    1.14257811990227776e-01, 8.75213679118525256e+00, -2.16929787526329321e+00,
    1.16210936696872255e-01, 8.60504207627572093e+00, -2.15234831939887172e+00,
    1.18164061975360682e-01, 8.46280995492959498e+00, -2.13568126444263484e+00,
    1.20117187499996322e-01, 8.32520325203277523e+00, -2.11928745022706622e+00,
    1.22070312499895098e-01, 8.19200000000703987e+00, -2.10315806829801133e+00,
    1.24023436774175100e-01, 8.06299217317146599e+00, -2.08728472499318229e+00,
    1.26953123746900931e-01, 7.87692315467275872e+00, -2.06393736501443570e+00,
    1.30859374098123454e-01, 7.64179109744297769e+00, -2.03363201254049386e+00,
    1.34765623780674720e-01, 7.42028992220936967e+00, -2.00421812948999545e+00,
    1.38671874242985771e-01, 7.21126764500034501e+00, -1.97564475345722457e+00,
    1.42578124148616536e-01, 7.01369867201821506e+00, -1.94786518986246371e+00,
    1.46484374166731490e-01, 6.82666670549979404e+00, -1.92083651719164372e+00,
    1.50390624434435488e-01, 6.64935067435644189e+00, -1.89451920694646070e+00,
    1.54296874339723084e-01, 6.48101268596180624e+00, -1.86887677685174936e+00,
    1.58203124999987427e-01, 6.32098765432149001e+00, -1.84387547036714849e+00,
    1.62109374999815342e-01, 6.16867469880220742e+00, -1.81948401724404896e+00,
    1.66015624243955634e-01, 6.02352943919619310e+00, -1.79567337310324682e+00,
    1.69921874302298687e-01, 5.88505749542848644e+00, -1.77241651049093640e+00,
    1.73828124315277527e-01, 5.75280901142480605e+00, -1.74968825924644555e+00,
    1.77734374286237506e-01, 5.62637364896854919e+00, -1.72746512253855222e+00,
    1.81640624146994889e-01, 5.50537636993989743e+00, -1.70572513658236602e+00,
    1.85546874316304788e-01, 5.38947370406942916e+00, -1.68444773712372431e+00,
    1.89453124405085355e-01, 5.27835053203882509e+00, -1.66361364967629299e+00,
    1.93359374570531595e-01, 5.17171718320401652e+00, -1.64320477712600699e+00,
    1.97265624263334577e-01, 5.06930694962380368e+00, -1.62320411193263148e+00,
    2.01171874086291030e-01, 4.97087380898513764e+00, -1.60359564135180399e+00,
    2.05078123979995308e-01, 4.87619050044336610e+00, -1.58436427985572159e+00,
    2.08984373896073439e-01, 4.78504675424820736e+00, -1.56549579585994181e+00,
    2.12890623963011144e-01, 4.69724772930228163e+00, -1.54697674768135762e+00,
    2.16796874723889421e-01, 4.61261261848719517e+00, -1.52879442500076479e+00,
    2.20703124198150608e-01, 4.53097346778917753e+00, -1.51093680996032553e+00,
    2.24609374375627030e-01, 4.45217392541970725e+00, -1.49339249945607477e+00,
    2.28515625000094036e-01, 4.37606837606657528e+00, -1.47615069024134016e+00,
    2.32421873924349737e-01, 4.30252102831546246e+00, -1.45920113655598627e+00,
    2.36328123935216378e-01, 4.23140497774241098e+00, -1.44253408394829741e+00,
    2.40234375000066919e-01, 4.16260162601510064e+00, -1.42614026966681173e+00,
    2.44140623863132178e-01, 4.09600001907347711e+00, -1.41001089239381727e+00,
    2.48046874999894917e-01, 4.03149606299383390e+00, -1.39413753858134015e+00,
    2.53906248590769879e-01, 3.93846156032078243e+00, -1.37079018013412401e+00,
    2.61718748558906533e-01, 3.82089554342693294e+00, -1.34048483059486401e+00,
    2.69531249159214337e-01, 3.71014493910979404e+00, -1.31107094300173976e+00,
    2.77343749428383191e-01, 3.60563381024826013e+00, -1.28249756949928795e+00,
    2.85156249289339359e-01, 3.50684932380819214e+00, -1.25471800582335113e+00,
    2.92968749999700462e-01, 3.41333333333682321e+00, -1.22768933094427446e+00,
    3.00781248554318814e-01, 3.32467534065511261e+00, -1.20137202743229921e+00,
    3.08593748521894806e-01, 3.24050634463533127e+00, -1.17572959680235023e+00,
    3.16406249999639899e-01, 3.16049382716409077e+00, -1.15072828980826181e+00,
    3.24218749999785061e-01, 3.08433734939963511e+00, -1.12633683668362750e+00,
    3.32031248841858584e-01, 3.01176471638753718e+00, -1.10252619147729547e+00,
    3.39843749265406558e-01, 2.94252874199264314e+00, -1.07926932798654107e+00,
    3.47656249999834799e-01, 2.87640449438338930e+00, -1.05654107474789782e+00,
    3.55468749999899247e-01, 2.81318681318761055e+00, -1.03431793796299587e+00,
    3.63281249999864997e-01, 2.75268817204403371e+00, -1.01257795132667816e+00,
    3.71093749064121570e-01, 2.69473684890124421e+00, -9.91300555400967731e-01,
    3.78906249999751032e-01, 2.63917525773369288e+00, -9.70466465976836723e-01,
    3.86718748879039009e-01, 2.58585859335407608e+00, -9.50057597243619156e-01,
    3.94531249999987899e-01, 2.53465346534661240e+00, -9.30056927638333697e-01,
    4.02343749999485523e-01, 2.48543689320706163e+00, -9.10448456251205407e-01,
    4.10156249578856991e-01, 2.43809524059864202e+00, -8.91217095348825872e-01,
    4.17968749447214571e-01, 2.39252336765021800e+00, -8.72348611340208357e-01,
    4.25781248601723117e-01, 2.34862386092395203e+00, -8.53829565534445223e-01,
    4.33593749393073047e-01, 2.30630630953458038e+00, -8.35647244566987801e-01,
    4.41406248572254134e-01, 2.26548673299152270e+00, -8.17789629001761220e-01,
    4.49218749348472501e-01, 2.22608695975035964e+00, -8.00245317566669279e-01,
    4.57031249277175089e-01, 2.18803419149470768e+00, -7.83003511263371976e-01,
    4.64843748529596368e-01, 2.15126051100659366e+00, -7.66053954531254355e-01,
    4.72656248830947701e-01, 2.11570248457175136e+00, -7.49386901356188240e-01,
    4.80468748609962581e-01, 2.08130081902951236e+00, -7.32993092000230995e-01,
    4.88281249241778237e-01, 2.04800000318021258e+00, -7.16863708730099525e-01,
    4.96093748931098810e-01, 2.01574803583926521e+00, -7.00990360175606675e-01,
    5.07812497779701388e-01, 1.96923077784079825e+00, -6.77642998396260410e-01,
    5.23437498033319737e-01, 1.91044776837204044e+00, -6.47337648285891021e-01,
    5.39062498006593560e-01, 1.85507247062801328e+00, -6.17923763020271188e-01,
    5.54687498964024250e-01, 1.80281690477552603e+00, -5.89350388745976339e-01,
    5.70312499806522322e-01, 1.75342465812909332e+00, -5.61570823110474571e-01,
    5.85937497921867001e-01, 1.70666667271966777e+00, -5.34542153929987052e-01,
    6.01562498226483444e-01, 1.66233766723853860e+00, -5.08224845014116688e-01,
    6.17187498682654212e-01, 1.62025316801528496e+00, -4.82582413587029357e-01,
    6.32812500000264566e-01, 1.58024691357958624e+00, -4.57581109246760320e-01,
    6.48437499353274216e-01, 1.54216867623689291e+00, -4.33189657120379490e-01,
    6.64062498728508976e-01, 1.50588235582451335e+00, -4.09379009344016609e-01,
    6.79687498865382267e-01, 1.47126437027210688e+00, -3.86122146934356092e-01,
    6.95312498728747119e-01, 1.43820224982050338e+00, -3.63393896015796081e-01,
    7.10937499999943157e-01, 1.40659340659351906e+00, -3.41170757402847080e-01,
    7.26562499999845568e-01, 1.37634408602179792e+00, -3.19430770766573779e-01,
    7.42187500000120126e-01, 1.34736842105241350e+00, -2.98153372318914478e-01,
    7.57812499999581890e-01, 1.31958762886670744e+00, -2.77319285416786077e-01,
    7.73437498602746576e-01, 1.29292929526503420e+00, -2.56910415591577124e-01,
    7.89062500000142664e-01, 1.26732673267303819e+00, -2.36909747078176913e-01,
    8.04687500000259015e-01, 1.24271844660154174e+00, -2.17301275689659512e-01,
    8.20312499999677036e-01, 1.21904761904809900e+00, -1.98069913762487504e-01,
    8.35937499999997113e-01, 1.19626168224299478e+00, -1.79201429457714445e-01,
    8.51562499999758749e-01, 1.17431192660583728e+00, -1.60682381690756770e-01,
    8.67187500000204725e-01, 1.15315315315288092e+00, -1.42500062607046951e-01,
    8.82812500000407896e-01, 1.13274336283133503e+00, -1.24642445206814556e-01,
    8.98437499999816813e-01, 1.11304347826109651e+00, -1.07098135556570995e-01,
    9.14062499999708455e-01, 1.09401709401744296e+00, -8.98563291221800009e-02,
    9.29687500000063949e-01, 1.07563025210076635e+00, -7.29067708080189947e-02,
    9.45312499999844014e-01, 1.05785123966959604e+00, -5.62397183230410880e-02,
    9.60937500000120459e-01, 1.04065040650393459e+00, -3.98459085470743157e-02,
    9.76562499999976685e-01, 1.02400000000002445e+00, -2.37165266173399170e-02,
    9.92187500000169420e-01, 1.00787401574785940e+00, -7.84317746085513856e-03,
    1.01562500000004907e+00, 9.84615384615337041e-01,  1.55041865360135717e-02,
    1.04687500000009237e+00, 9.55223880596930641e-01,  4.58095360313824362e-02,
    1.07812500000002154e+00, 9.27536231884039442e-01,  7.52234212376075018e-02,
    1.10937499999982481e+00, 9.01408450704367703e-01,  1.03796793681485644e-01,
    1.14062500000007416e+00, 8.76712328767066285e-01,  1.31576357788784293e-01,
    1.17187500000009659e+00, 8.53333333333263000e-01,  1.58605030176721007e-01,
    1.20312499999950173e+00, 8.31168831169175393e-01,  1.84922338493597849e-01,
    1.23437500000022027e+00, 8.10126582278336449e-01,  2.10564769107528083e-01,
    1.26562500000064615e+00, 7.90123456789720069e-01,  2.35566071313277448e-01,
    1.29687500000144706e+00, 7.71084337348537208e-01,  2.59957524438041876e-01,
    1.32812499999945932e+00, 7.52941176470894757e-01,  2.83768173130237500e-01,
    1.35937500055846350e+00, 7.35632183605830825e-01,  3.07025035705735583e-01,
    1.39062499999999467e+00, 7.19101123595508374e-01,  3.29753286372464149e-01,
    1.42187500000017564e+00, 7.03296703296616421e-01,  3.51976423157301710e-01,
    1.45312500161088876e+00, 6.88172042247866766e-01,  3.73716410902152685e-01,
    1.48437500134602307e+00, 6.73684209915422660e-01,  3.94993809147663466e-01,
    1.51562499999932343e+00, 6.59793814433284220e-01,  4.15827895143264570e-01,
    1.54687500000028200e+00, 6.46464646464528614e-01,  4.36236766775100371e-01,
    1.57812500000061906e+00, 6.33663366336385092e-01,  4.56237433481979870e-01,
    1.60937500243255216e+00, 6.21359222361793417e-01,  4.75845906381452632e-01,
    1.64062500000026312e+00, 6.09523809523711768e-01,  4.95077266798011895e-01,
    1.67187500000027911e+00, 5.98130841121395473e-01,  5.13945751102401260e-01,
    1.70312500224662178e+00, 5.87155962528224662e-01,  5.32464800188589216e-01,
    1.73437500283893620e+00, 5.76576575632799071e-01,  5.50647119589526390e-01,
    1.76562500399259092e+00, 5.66371680135198341e-01,  5.68504737613959144e-01,
    1.79687500443862880e+00, 5.56521737755718449e-01,  5.86049047473771623e-01,
    1.82812500114411280e+00, 5.47008546666207462e-01,  6.03290852063923744e-01,
    1.85937500250667465e+00, 5.37815125325376786e-01,  6.20240411099985067e-01,
    1.89062500504214515e+00, 5.28925618424108568e-01,  6.36907464903988974e-01,
    1.92187500371610143e+00, 5.20325202245941476e-01,  6.53301273946326866e-01,
    1.95312500494870611e+00, 5.11999998702726389e-01,  6.69430656476366792e-01,
    1.98437500351688123e+00, 5.03937006980894941e-01,  6.85304004871206018e-01,
    2.03125000000003997e+00, 4.92307692307682621e-01,  7.08651367095930240e-01,
    2.09375000579615866e+00, 4.77611938976327366e-01,  7.38956719359554093e-01,
    2.15625000000061062e+00, 4.63768115941897652e-01,  7.68370601797816022e-01,
    2.21875000323311955e+00, 4.50704224695355204e-01,  7.96943975698769513e-01,
    2.28125000853738547e+00, 4.38356162743050726e-01,  8.24723542091080120e-01,
    2.34374999999916556e+00, 4.26666666666818573e-01,  8.51752210736227866e-01,
    2.40625000438447856e+00, 4.15584414827170512e-01,  8.78069520876078258e-01,
    2.46875000884389584e+00, 4.05063289688167072e-01,  9.03711953249632494e-01,
    2.53124999999940403e+00, 3.95061728395154743e-01,  9.28713251872476775e-01,
    2.59375000434366632e+00, 3.85542168029044230e-01,  9.53104706671537905e-01,
    2.65625000734081196e+00, 3.76470587194880080e-01,  9.76915356454189698e-01,
    2.71875000787161980e+00, 3.67816090889081959e-01,  1.00017221875016560e+00,
    2.78125001557333462e+00, 3.59550559784484969e-01,  1.02290047253181449e+00,
    2.84375001147093220e+00, 3.51648350229895601e-01,  1.04512360775085789e+00,
    2.90625000771072894e+00, 3.44086020592463127e-01,  1.06686359300668343e+00,
    2.96875001371853831e+00, 3.36842103706616824e-01,  1.08814099342179560e+00,
    3.03125000512624965e+00, 3.29896906658595002e-01,  1.10897507739479018e+00,
    3.09375001373132807e+00, 3.23232321797685962e-01,  1.12938395177327244e+00,
    3.15625001204422961e+00, 3.16831681959289180e-01,  1.14938461785752644e+00,
    3.21875000888250318e+00, 3.10679610793130057e-01,  1.16899308818952186e+00,
    3.28125000000102052e+00, 3.04761904761809976e-01,  1.18822444735810784e+00,
    3.34375001587649123e+00, 2.99065419140752298e-01,  1.20709293641028914e+00,
    3.40625000791328070e+00, 2.93577980969346064e-01,  1.22561198175258212e+00,
    3.46875000615970519e+00, 2.88288287776354346e-01,  1.24379430028837845e+00,
    3.53125000516822674e+00, 2.83185840293502689e-01,  1.26165191737618265e+00,
    3.59375001425228779e+00, 2.78260868461675415e-01,  1.27919622952937750e+00,
    3.65625001719730669e+00, 2.73504272217836075e-01,  1.29643803670156643e+00,
    3.71875000856489324e+00, 2.68907562405871714e-01,  1.31338759261496740e+00,
    3.78125001788371806e+00, 2.64462808666557803e-01,  1.33005464752659286e+00,
    3.84375001532508964e+00, 2.60162600588744020e-01,  1.34644845655970613e+00,
    3.90625000429340918e+00, 2.55999999718627136e-01,  1.36257783560168733e+00,
    3.96875001912740766e+00, 2.51968502722644594e-01,  1.37845118847836900e+00,
    4.06250002536431332e+00, 2.46153844616978895e-01,  1.40179855389937913e+00,
    4.18750001743208244e+00, 2.38805969155131859e-01,  1.43210390131407017e+00,
    4.31250002253733200e+00, 2.31884056759177282e-01,  1.46151778758352613e+00,
    4.43750000671406397e+00, 2.25352112335092170e-01,  1.49009115631456268e+00,
    4.56250002627485340e+00, 2.19178080929562313e-01,  1.51787072466748185e+00,
    4.68750001185115028e+00, 2.13333332793974317e-01,  1.54489939382477459e+00,
    4.81250001682742301e+00, 2.07792207065640028e-01,  1.57121670311050998e+00,
    4.93750000000042366e+00, 2.02531645569602875e-01,  1.59685913022732606e+00,
    5.06249999999927613e+00, 1.97530864197559108e-01,  1.62186043243251454e+00,
    5.18750002327641901e+00, 1.92771083472381588e-01,  1.64625189004383721e+00,
    5.31250002381002329e+00, 1.88235293273997795e-01,  1.67006253873242194e+00,
    5.43750000000577405e+00, 1.83908045976816203e-01,  1.69331939641586438e+00,
    5.56250002193114934e+00, 1.79775280190080267e-01,  1.71604765143503712e+00,
    5.68749999999938005e+00, 1.75824175824194989e-01,  1.73827078427695980e+00,
    5.81250002749782002e+00, 1.72043009938785768e-01,  1.76001077564428243e+00,
    5.93749999999874767e+00, 1.68421052631614471e-01,  1.78128816936054868e+00,
    6.06250001966917473e+00, 1.64948453073088669e-01,  1.80212225950800153e+00,
    6.18750003004243609e+00, 1.61616160831459688e-01,  1.82253113275015188e+00,
    6.31250002448351388e+00, 1.58415840969730465e-01,  1.84253179848005466e+00,
    6.43750001359968849e+00, 1.55339805497076044e-01,  1.86214026810242750e+00,
    6.56250003345742350e+00, 1.52380951604072529e-01,  1.88137163301601618e+00,
    6.68750002403557531e+00, 1.49532709742937614e-01,  1.90024011581622965e+00,
    6.81250003423489581e+00, 1.46788990088028509e-01,  1.91875916501466826e+00,
    6.93750003062940923e+00, 1.44144143507740546e-01,  1.93694148348760287e+00,
    7.06250002747386052e+00, 1.41592919803171097e-01,  1.95479910036266347e+00,
    7.18750003617887856e+00, 1.39130434082284093e-01,  1.97234341115705192e+00,
    7.31250000000050537e+00, 1.36752136752127301e-01,  1.98958521255804399e+00,
    7.43750002212249761e+00, 1.34453781112678528e-01,  2.00653477384620160e+00,
    7.56250003604752941e+00, 1.32231404328381430e-01,  2.02320182812357530e+00,
    7.68750005007207449e+00, 1.30081299965731312e-01,  2.03959563964607682e+00,
    7.81249996125652668e+00, 1.28000000634773070e-01,  2.05572501010335529e+00,
    7.93750005224239974e+00, 1.25984251139310915e-01,  2.07159837080052966e+00,
    8.12500004244456164e+00, 1.23076922433975874e-01,  2.09494573343974722e+00,
    8.37500006149772425e+00, 1.19402984197849338e-01,  2.12525108505414195e+00,
    8.62500006593247370e+00, 1.15942028099206410e-01,  2.15466497056176820e+00,
    8.87500007743793873e+00, 1.12676055354884341e-01,  2.18323834408688100e+00,
    9.12500001754142609e+00, 1.09589040885222130e-01,  2.21101790139090326e+00,
    9.37500007707016181e+00, 1.06666665789779500e-01,  2.23804658007729174e+00,
    9.62500004426353151e+00, 1.03896103418305616e-01,  2.26436388477265638e+00,
    9.87500006518495788e+00, 1.01265822116353585e-01,  2.29000631738819393e+00,
    1.01250000000026539e+01, 9.87654320987395445e-02,  2.31500761299286495e+00,
    1.03750000409819823e+01, 9.63855417879450060e-02,  2.33939907006683256e+00,
    1.06250000362555337e+01, 9.41176467376672460e-02,  2.36320971822276604e+00,
    1.08750000879032314e+01, 9.19540222452362582e-02,  2.38646658505780351e+00,
    1.11250000697274576e+01, 8.98876398860551373e-02,  2.40919483431994053e+00,
    1.13750000462194141e+01, 8.79120875548795450e-02,  2.43141796890025930e+00,
    1.16250000714972366e+01, 8.60215048472860316e-02,  2.45315795762371991e+00,
    1.18750000788855150e+01, 8.42105257563797310e-02,  2.47443535656369562e+00,
    1.21250000895724916e+01, 8.24742261948517991e-02,  2.49526944421096886e+00,
    1.23750000985058719e+01, 8.08080801648427965e-02,  2.51567831641482442e+00,
    1.26250000894226950e+01, 7.92079202310506381e-02,  2.53567898224440924e+00,
    1.28750000768594433e+01, 7.76699024489580225e-02,  2.55528745251946532e+00,
    1.31250000578007420e+01, 7.61904758549435401e-02,  2.57451881288155349e+00,
    1.33750000809310077e+01, 7.47663546877819496e-02,  2.59338729883298669e+00,
    1.36250000915049636e+01, 7.33944949199294983e-02,  2.61190634726526838e+00,
    1.38750000830616607e+01, 7.20720716406179490e-02,  2.63008866561892418e+00,
    1.41249999999960103e+01, 7.07964601770111474e-02,  2.64794627703222218e+00,
    1.43750000290097564e+01, 6.95652172509168693e-02,  2.66549058870148414e+00,
    1.46250000868097665e+01, 6.83760679702078294e-02,  2.68273239905363070e+00,
    1.48750000966053975e+01, 6.72268903196987927e-02,  2.69968195792617394e+00,
    1.51250001097012756e+01, 6.61157019998031836e-02,  2.71634901116988203e+00,
    1.53750000510427132e+01, 6.50406501905787804e-02,  2.73274281701243282e+00,
    1.56250001080665442e+01, 6.39999995573594382e-02,  2.74887220253872400e+00,
    1.58750000434989929e+01, 6.29921258116476201e-02,  2.76474554751884938e+00,
    1.62500000641781739e+01, 6.15384612954199342e-02,  2.78809291272517257e+00,
    1.67500001015987401e+01, 5.97014921751882754e-02,  2.81839826433667184e+00,
    1.72500001048300184e+01, 5.79710141404578272e-02,  2.84781214955447126e+00,
    1.77500001262529885e+01, 5.63380277682904579e-02,  2.87638552303426920e+00,
    1.82500001543340602e+01, 5.47945200845665337e-02,  2.90416508848516131e+00,
    1.87500001096404212e+01, 5.33333330214672482e-02,  2.93119375826390893e+00,
    1.92500001680268191e+01, 5.19480514946147609e-02,  2.95751106946245912e+00,
    1.97500000329124035e+01, 5.06329113080278073e-02,  2.98315349301358168e+00,
    2.02500001270002485e+01, 4.93827157396732261e-02,  3.00815479982416534e+00,
    2.07500001519906796e+01, 4.81927707313324349e-02,  3.03254625400155930e+00,
    2.12500001425219267e+01, 4.70588232137922752e-02,  3.05635690207734001e+00,
    2.17500000758314478e+01, 4.59770113339538697e-02,  3.07961376102119644e+00,
    2.22500001767207358e+01, 4.49438198677525880e-02,  3.10234201655475417e+00,
    2.27500001365873317e+01, 4.39560436921389575e-02,  3.12456515140079816e+00,
    2.32500001697599998e+01, 4.30107523741288036e-02,  3.14630513933487066e+00,
    2.37500001766865303e+01, 4.21052628446554611e-02,  3.16758253792008304e+00,
  };
  
#define        HIWORD                1
#define        LOWORD                0
  
  // Fast Log - Stripped version of libm's Log(x)
  /* This is the libm log stripped down with slightly less precision and no handling for inf, 0 etc.
     For an idea of how this works see: http://www.informatik.uni-trier.de/Reports/TR-08-2004/rnc6_07_dinechin.pdf
     This is about 15% faster than the standard log() */
  inline double fastLog(const double x)
  {
    union
    {
        int    i[2];
        double x;
    } u;
    
    u.x = x;
    
    /* get exponent part, store in dn */
    const int hx = u.i[HIWORD];     // Get the top half of the double
    const int ix = hx & 0x7fffffff; // condition ix
    double dn    = (double)((ix >> 20) - 0x3ff); // Shift to get E
    
    /* Now we take advantage of the float point identity:
       
       log(x) = E × log(2) + log(y)
       
       Where E is the exponent part of the float and y is from the mentisa.
       
       y is then set to be on the interval from 11/16 to 23/16 using
       a bit shift to divide it by 2.
       
       log(y) needs to be scaled and approximated on a table. Then we use the
       Horner scheme polynomial to approximated the final log */
    
    const double dn1 = dn * LN2HI;                       // Low Order E * log(2)
    int i            = (ix & 0x000fffff) | 0x3ff00000;   // scale x to [1,2]
    u.i[HIWORD]      = i;                                // copy back E
    i                = (i - 0x3fb80000) >> 15;           // mantisa table addr
    const double *tb = (double *)_TBL_log + (i + i + i); // Get table const's
    const double s   = (u.x - tb[0]) * tb[1];            // initial estimate
    dn               = dn * LN2LO + tb[2];               // High order E * log(2)
    
    // 8th order Polynomial in Horner form
    return (dn1 + (dn + ((B1 * s) * (B2 + s * (B3 + s))) *
                   (((B4 + s * B5) + (s * s) * (B6 + s)) *
                    (B7 + s * (B8 + s)))));
    
  };
 
  /* data for a Chebyshev series over a given interval */
  struct cheb_series_struct {
      double * c;   /* coefficients                */
      int order;    /* order of expansion          */
      double a;     /* lower interval point        */
      double b;     /* upper interval point        */
      int order_sp; /* effective single precision order */
  };
  typedef struct cheb_series_struct cheb_series;
  
  /* Chebyshev fits from SLATEC code for psi(x) */
  static double psics_data[23] = {
    -.038057080835217922,
    .491415393029387130,
    -.056815747821244730,
    .008357821225914313,
    -.001333232857994342,
    .000220313287069308,
    -.000037040238178456,
    .000006283793654854,
    -.000001071263908506,
    .000000183128394654,
    -.000000031353509361,
    .000000005372808776,
    -.000000000921168141,
    .000000000157981265,
    -.000000000027098646,
    .000000000004648722,
    -.000000000000797527,
    .000000000000136827,
    -.000000000000023475,
    .000000000000004027,
    -.000000000000000691,
    .000000000000000118,
    -.000000000000000020
  };
  const static cheb_series psi_cs = { psics_data, 22, -1, 1, 17 };
  
  static double apsics_data[16] = {
    -.0204749044678185,
    -.0101801271534859,
    .0000559718725387,
    -.0000012917176570,
    .0000000572858606,
    -.0000000038213539,
    .0000000003397434,
    -.0000000000374838,
    .0000000000048990,
    -.0000000000007344,
    .0000000000001233,
    -.0000000000000228,
    .0000000000000045,
    -.0000000000000009,
    .0000000000000002,
    -.0000000000000000
  };
  const static cheb_series apsi_cs = { apsics_data, 15, -1, 1, 9 };
 
  inline double cheb_eval_psi(const double x)
  {
    double d  = 0.0, dd = 0.0;
    
    const double y2 = 2.0 * x;
    
    for (uint j = 22; j >= 1; j--)
    {
      const double temp = d;
      d = y2*d - dd + psics_data[j];
      dd = temp;
    }
    
    return x*d - dd + 0.5 * -.038057080835217922;
  }
  
  inline double cheb_eval_apsi(const double x)
  {
    double d  = 0.0, dd = 0.0;
    
    const double y2 = 2.0 * x;
    
    for (uint j = 15; j >= 1; j--)
    {
      const double temp = d;
      d = y2*d - dd + apsics_data[j];
      dd = temp;
    }
    
    return x*d - dd + 0.5 * -.0204749044678185;
  }
 
 
  
  inline double fast_psi(double x)
  {
    if (x >= 2.0)
    {
      const double t    = 8.0 / (x * x) - 1.0;
      const double cheb = cheb_eval_apsi(t);
      return fastLog(x) - 0.5 / x + cheb;
    }
    else if (x < 1.0)
    {
      const double t1   = 1.0 / x;
      const double cheb = cheb_eval_psi(2.0 * x - 1.0);
      return cheb - t1;
    }
    else
    {
      const double v = x - 1.0;
      return cheb_eval_psi(2.0 * v - 1.0);
    }
  }
 
  // Natural Log of 2 DEFINED in cmath as M_LN2, but is not 128 bit
#define D_LOG_2            0.6931471805599453094172321214581766
 
  // Convert a number into Wows, for instance a KL distance
  /* Wow's in this computation is the same as the conversion from Nats to Bits in KL */
  template <class T> inline T wow(const T& k)
  {
    // surprise should always be positive but rounding may mess that up: This is OK since surprise is negative with
    // equality if and only if two distribution functions are equal. Thus, we gain no insite with negative surprise.
    
    if (k < 0.0) return 0.0;
    // return the surprise, in units of wows (KL Bits):
    return k * (1.0F / D_LOG_2);
  }
 
  // Compute the KL distance for a single gamma PDF
  template <class T> inline T KLgamma(const T a, const T b,
                                      const T A, const T B,
                                      const bool doWow = false)
  {
    /* We will compute the joint KL for P(x), P(X), P(y) and P(Y) where one is a gamma distribution and the other
       gaussian as:
 
       surprise is KL(new || old):
 
                               P(X)
       KL = Integrate[P(x) log[----] ]
                               P(x)
 
       a - New Alpha
       b - New Beta
       A - Current Alpha
       B - Current Beta
 
                    a B         b        Gamma[A]
       KL    = -a + --- + A Log[-] + Log[--------] + a PolyGamma[0, a] - A * PolyGamma[0, a]
                     b          B        Gamma[a]
 
       Note: Since the pure Gamma function can get large very quickly, we avoid large numbers by computing the log gamma
       directly as lgamma. */
 
    T const k =  - a + a * B / b + A * fastLog(b / B) + lgamma(A) - lgamma(a) + (a - A) * fast_psi(a);
 
    if (doWow) return wow(k); else return k;
  }
}
 
// ##############################################################################################################
double Surprise::process(jevois::RawImage const & input)
{
  std::string const chans = channels::get();
 
  // Compute feature maps and saliency maps, possibly gist. Results are stored in the Saliency class:
  itsSaliency->process(input, (chans.find('G') != chans.npos));
 
  // Aggregate our data values from all maps. These maps are small, no need to parallelize:
  std::vector<float> data; std::string done;
 
  for (char c : chans)
  {
    if (done.find(c) != done.npos) continue; // skip duplicates
    done += c; // mark this channel as done
    intg32 * pix; size_t siz;
 
    switch (c)
    {
    case 'S': pix = itsSaliency->salmap.pixels; siz = env_img_size(&itsSaliency->salmap); break;
    case 'I': pix = itsSaliency->intens.pixels; siz = env_img_size(&itsSaliency->intens); break;
    case 'C': pix = itsSaliency->color.pixels; siz = env_img_size(&itsSaliency->color); break;
    case 'O': pix = itsSaliency->ori.pixels; siz = env_img_size(&itsSaliency->ori); break;
    case 'F': pix = itsSaliency->flicker.pixels; siz = env_img_size(&itsSaliency->flicker); break;
    case 'M': pix = itsSaliency->motion.pixels; siz = env_img_size(&itsSaliency->motion); break;
    case 'G':
    {
      unsigned char const * g = itsSaliency->gist;
      for (size_t i = 0; i < itsSaliency->gist_size; ++i) data.push_back(g[i]);
      continue;
    }
    default: continue; // should never happen given our regex spec for the parameter
    }
 
    // Concatenate the data if it was not gist:
    for (size_t i = 0; i < siz; ++i) data.push_back(pix[i]);
  }
  
  size_t const datasiz = data.size(); // final data size
  float const ufac = updatefac::get(); // get() is somewhat expensive (requires mutex lock), so cache it here.
  float const initfac = 1.0F / (1.0F - ufac);
  
  // Initialize the prior if this is our first frame, or frame size or map size just changed somehow. We initialize
  // alpha and beta as in the SurpriseModelSP of the iLab Neuromorphic C++ Vision Toolkit, from which this
  // implementation is derived. Also see Itti & Baldi, Vis Res, 2009, for details:
 
  if (itsAlpha.size() != datasiz)
  {
    itsAlpha.clear(); itsBeta.clear();
    for (float d : data)
    {
      itsAlpha.push_back(d * initfac);
      itsBeta.push_back(initfac);
    }
  }
    
  // Compute posterior and KL, independently for every entry in our vectors. Here we assume Poisson data and a Gamma
  // conjugate prior, as in Itti & Baldi, Vision Research, 2009. Note: we compute surprise using double precision, then
  // store alpha and beta using float:
  double surprise = 0.0;
 
  for (size_t i = 0; i < datasiz; ++i)
  {
    // First, decay alpha and beta. Make sure alpha does not decay all the way to 0:
    double alpha = itsAlpha[i] * ufac, beta = itsBeta[i] * ufac;
    if (alpha < 1.0e-5) alpha = 1.0e-5;
 
    // Compute the posterior:
    double const newAlpha = alpha + data[i];
    double const newBeta  = beta  + 1.0F;
 
    // Surprise is KL(new || old). Keep track of the max value found over the data array:
    double const s = std::abs(KLgamma<double>(newAlpha, newBeta, alpha, beta, true));
    if (s > surprise) surprise = s;
    
    // The posterior becomes our new prior for the next video frame:
    itsAlpha[i] = newAlpha; itsBeta[i] = newBeta;
  }
 
  // Return max number of wows found over the whole data array:
  return surprise;
}