Nav apraksta

Alexander Freytag bcf596b43c separated Matlab-NICE-Conversion from GPHIK-Matlab-Interface 11 gadi atpakaļ
algebra 72d211d8f0 init 12 gadi atpakaļ
configs 72d211d8f0 init 12 gadi atpakaļ
doxy 72d211d8f0 init 12 gadi atpakaļ
kernels 72d211d8f0 init 12 gadi atpakaļ
matlab bcf596b43c separated Matlab-NICE-Conversion from GPHIK-Matlab-Interface 11 gadi atpakaļ
parameterizedFunctions 7fc06c0490 removed compiler warnings 11 gadi atpakaļ
progs 7fc06c0490 removed compiler warnings 11 gadi atpakaļ
tests b0539f5af2 fixed bug for using prev alphas in IL settings, better code consistency, clean-up, more testing 11 gadi atpakaļ
tutorial 72d211d8f0 init 12 gadi atpakaļ
FMKGPHyperparameterOptimization.cpp dce3fe031a consistency changes for incremental update with differenct variance approximations 11 gadi atpakaļ
FMKGPHyperparameterOptimization.h d755cd9501 better doxygen docu for main classes 11 gadi atpakaļ
FastMinKernel.cpp 4158390858 removed merging issue with previous official release 11 gadi atpakaļ
FastMinKernel.h afb7a2c6a0 GPHIK IL - more test units, running IL on empty classifier is now possible, updated addMultipleExamples 11 gadi atpakaļ
FeatureMatrixT.h afb7a2c6a0 GPHIK IL - more test units, running IL on empty classifier is now possible, updated addMultipleExamples 11 gadi atpakaļ
FeatureMatrixT.tcc afb7a2c6a0 GPHIK IL - more test units, running IL on empty classifier is now possible, updated addMultipleExamples 11 gadi atpakaļ
GMHIKernel.cpp e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
GMHIKernel.h e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
GPHIKClassifier.cpp 9a2118674b security check in train method added 11 gadi atpakaļ
GPHIKClassifier.h d755cd9501 better doxygen docu for main classes 11 gadi atpakaļ
GPLikelihoodApprox.cpp b0539f5af2 fixed bug for using prev alphas in IL settings, better code consistency, clean-up, more testing 11 gadi atpakaļ
GPLikelihoodApprox.h e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
IKMLinearCombination.cpp 7fc06c0490 removed compiler warnings 11 gadi atpakaļ
IKMLinearCombination.h e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
IKMNoise.cpp e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
IKMNoise.h e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
ImplicitKernelMatrix.cpp 72d211d8f0 init 12 gadi atpakaļ
ImplicitKernelMatrix.h e26c33948b OnlineLearnable added, GPHIK ready for IL with 1 example update, everything nice and clean 11 gadi atpakaļ
License 72d211d8f0 init 12 gadi atpakaļ
Makefile 72d211d8f0 init 12 gadi atpakaļ
Makefile.inc 72d211d8f0 init 12 gadi atpakaļ
OnlineLearnable.h d755cd9501 better doxygen docu for main classes 11 gadi atpakaļ
Quantization.cpp 72d211d8f0 init 12 gadi atpakaļ
Quantization.h 72d211d8f0 init 12 gadi atpakaļ
README 72d211d8f0 init 12 gadi atpakaļ
SortedVectorSparse.h b7eb0a2202 Persisent in new layout for main classes of gp-hik-core implemented 11 gadi atpakaļ
VectorSorter.cpp 72d211d8f0 init 12 gadi atpakaļ
VectorSorter.h 72d211d8f0 init 12 gadi atpakaļ
doxyfile_gp_hik.txt 72d211d8f0 init 12 gadi atpakaļ
libdepend.inc 72d211d8f0 init 12 gadi atpakaļ
tools.h 72d211d8f0 init 12 gadi atpakaļ

README

This code requires the NICE-core library (available on github) and implements the ideas presented in the following papers:

Alexander Freytag and Erik Rodner and Paul Bodesheim and Joachim Denzler
"Rapid Uncertainty Computation with Gaussian Processes and Histogram Intersection Kernels"
Asian Conference on Computer Vision (ACCV). 2012.

Erik Rodner and Alexander Freytag and Paul Bodesheim and Joachim Denzler
"Large-Scale Gaussian Process Classification with Flexible Adaptive Histogram Kernels"
European Conference on Computer Vision (ECCV). 2012. pages 85--98

If you use this software in your research, you have to cite at least one of the two papers.

Problems, questions, and general feedback should be addressed to:

Alexander.Freytag uni-jena.de
Erik.Rodner uni-jena.de

Copyright (LGPL) Alexander Freytag and Erik Rodner