Initial import of code.

da2c1226 · lucasb-eyer · 3ae01ffb · da2c1226 · da2c1226 · da2c1226
Commit da2c1226 authored Nov 13, 2015 by lucasb-eyer
20 changed files
--- a/README.md
+++ b/README.md
-# pydensecrf
-Python wrapper to Philipp Krähenbühl's dense (fully connected) CRFs with gaussian edge potentials.
+PyDenseCRF
+==========
+
+This is a (Cython-based) Python wrapper for Philipp Krähenbühl's Fully-Connected CRFs (version 2).
+
+If you use this code for your reasearch, please cite:
+
+```
+Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
+Philipp Krähenbühl and Vladlen Koltun
+NIPS 2011
+```
+
+and provide a link to this repository as a footnote or a citation.
+
+Installation
+============
+
+You can install this using `pip` by executing:
+
+```
+TODO
+```
+
+Usage
+=====
+
+For images, the easiest way to use this library is using the `DenseCRF2D` class:
+
+```
+import numpy as np
+import densecrf as dcrf
+
+d = dcrf.DenseCRF2D(640, 480, 3)  # width, height, nlabels
+```
+
+Unary potential
+---------------
+
+You can then set a fixed unary potential in the following way:
+
+```
+U = np.array(...)     # Get the unary in some way.
+print(U.shape)        # -> (640, 480, 3)
+print(U.dtype)        # -> dtype('float32')
+U = U.reshape((-1,3)) # Needs to be flat.
+d.setUnaryEnergy(U)
+
+# Or alternatively: d.setUnary(ConstUnary(U))
+```
+
+Remember that `U` should be negative log-probabilities, so if you're using
+probabilities `py`, don't forget to `U = -np.log(py)` them.
+
+Requiring the `reshape` on the unary is an API wart that I'd like to fix, but
+don't know how to without introducing an explicit dependency on numpy.
+
+Pairwise potentials
+-------------------
+
+The two-dimensional case has two utility methods for adding the most-common pairwise potentials:
+
+```
+# This adds the color-independent term, features are the locations only.
+d.addPairwiseGaussian(sxy=(3,3), compat=3, kernel=dcrf.DIAG_KERNEL, normalization=dcrf.NORMALIZE_SYMMETRIC)
+
+# This adds the color-dependent term, i.e. features are (x,y,r,g,b).
+# im is an image-array, e.g. im.dtype == np.uint8 and im.shape == (640,480,3)
+d.addPairwiseBilateral(sxy=(80,80), srgb=(13,13,13), rgbim=im, compat=10, kernel=dcrf.DIAG_KERNEL, normalization=dcrf.NORMALIZE_SYMMETRIC)
+```
+
+Both of these methods have shortcuts and default-arguments such that the most
+common use-case can be simplified to:
+
+```
+d.addPairwiseGaussian(sxy=3, compat=3)
+d.addPairwiseBilateral(sxy=80, srgb=13, rgbim=im, compat=10)
+```
+
+### Compatibilities
+
+The `compatibility` argument can be any of the following:
+
+- A number, then a `PottsCompatibility` is being used.
+- A 1D array, then a `DiagonalCompatibility` is being used.
+- A 2D array, then a `MatrixCompatibility` is being used.
+
+### Kernels
+
+Possible values for the `kernel` argument are:
+
+- `CONST_KERNEL`
+- `DIAG_KERNEL` (the default)
+- `FULL_KERNEL`
+
+### Normalizations
+
+Possible values for the `normalization` argument are:
+
+- `NO_NORMALIZATION`
+- `NORMALIZE_BEFORE`
+- `NORMALIZE_AFTER`
+- `NORMALIZE_SYMMETRIC` (the default)
+
+Inference
+---------
+
+The easiest way to do inference is to simply call:
+
+```
+Q = d.inference(n_iterations=5)
+```
+
+And the MAP prediction is then:
+
+```
+map = np.argmax(Q, axis=0).reshape((640,480))
+```
+
+Step-by-step inference
+----------------------
+
+If for some reason you want to run the inference loop manually, you can do so:
+
+```
+Q, tmp1, tmp2 = d.startInference()
+for i in range(5):
+    print("KL-divergence at {}: {}".format(i, d.klDivergence(Q)))
+    d.stepInference(Q, tmp1, tmp2)
+```
+
+Generic non-2D
+--------------
+
+The `DenseCRF` class can be used for generic (non-2D) dense CRFs.
+Its usage is exactly the same as above, except that the 2D-specific pairwise
+potentials `addPairwiseGaussian` and `addPairwiseBilateral` are missing.
+
+Instead, you need to use the generic `addPairwiseEnergy` method like this:
+
+```
+d = dcrf.DenseCRF(100, 3)  # npoints, nlabels
+
+feats = np.array(...)  # Get the pairwise features from somewhere.
+print(feats.shape)     # -> (100, 3)
+print(feats.dtype)     # -> dtype('float32')
+
+dcrf.addPairwiseEnergy(feats)
+```
+
+In addition, you can pass `compatibility`, `kernel` and `normalization`
+arguments just like in the 2D gaussian and bilateral cases.
+
+The potential will be computed as `w*exp(-0.5 * |f_i - f_j|^2)`.
+
+Learning
+--------
+
+The learning has not been fully wrapped. If you need it, get in touch or better
+yet, wrap it and submit a pull-request!
--- a/densecrf.pxd
+++ b/densecrf.pxd
+from eigen cimport *
+
+
+cdef extern from "densecrf/include/labelcompatibility.h":
+    cdef cppclass LabelCompatibility:
+        pass
+
+    cdef cppclass PottsCompatibility(LabelCompatibility):
+        PottsCompatibility(float) except +
+
+    cdef cppclass DiagonalCompatibility(LabelCompatibility):
+        DiagonalCompatibility(const c_VectorXf&) except +
+
+    cdef cppclass MatrixCompatibility(LabelCompatibility):
+        MatrixCompatibility(const c_MatrixXf&) except +
+
+
+cdef extern from "densecrf/include/unary.h":
+    cdef cppclass UnaryEnergy:
+        pass
+
+    cdef cppclass ConstUnaryEnergy(UnaryEnergy):
+        ConstUnaryEnergy(const c_MatrixXf& unary) except +
+
+    cdef cppclass LogisticUnaryEnergy(UnaryEnergy):
+        LogisticUnaryEnergy(const c_MatrixXf& L, const c_MatrixXf& feature) except +
+
+
+cdef class Unary:
+    cdef UnaryEnergy *thisptr
+    cdef UnaryEnergy* move(self)
+
+
+cdef class ConstUnary(Unary):
+    pass
+
+
+cdef class LogisticUnary(Unary):
+    pass
+
+
+cdef extern from "densecrf/include/pairwise.h":
+    cpdef enum NormalizationType: NO_NORMALIZATION, NORMALIZE_BEFORE, NORMALIZE_AFTER, NORMALIZE_SYMMETRIC
+    cpdef enum KernelType: CONST_KERNEL, DIAG_KERNEL, FULL_KERNEL
+
+
+cdef extern from "densecrf/include/densecrf.h":
+    cdef cppclass c_DenseCRF "DenseCRF":
+        c_DenseCRF(int N, int M) except +
+
+        # Setup methods.
+        # TODO
+        #void addPairwiseEnergy(PairwisePotential *potential)
+        void addPairwiseEnergy(const c_MatrixXf &features, LabelCompatibility*, KernelType, NormalizationType)
+        void setUnaryEnergy(UnaryEnergy *unary)
+        void setUnaryEnergy(const c_MatrixXf &unary)
+        void setUnaryEnergy(const c_MatrixXf &L, const c_MatrixXf &feature)
+
+        # Inference methods.
+        c_MatrixXf inference(int n_iterations)
+        # TODO: Not enabled because it would require wrapping VectorXs (note the `s`)
+        #c_VectorXs map(int n_iterations)
+
+        # Step-by-step inference methods.
+        c_MatrixXf startInference() const
+        void stepInference(c_MatrixXf &Q, c_MatrixXf &tmp1, c_MatrixXf &tmp2) const
+        #double gradient( int n_iterations, const ObjectiveFunction & objective, c_VectorXf * unary_grad, c_VectorXf * lbl_cmp_grad, c_VectorXf * kernel_grad=NULL ) const;
+
+        double klDivergence(const c_MatrixXf &Q) const
+
+        #c_VectorXf unaryParameters() const;
+        #void setUnaryParameters( const c_VectorXf & v );
+        #c_VectorXf labelCompatibilityParameters() const;
+        #void setLabelCompatibilityParameters( const c_VectorXf & v );
+        #c_VectorXf kernelParameters() const;
+        #void setKernelParameters( const c_VectorXf & v );
+
+
+cdef extern from "densecrf/include/densecrf.h":
+    cdef cppclass c_DenseCRF2D "DenseCRF2D" (c_DenseCRF):
+        c_DenseCRF2D(int W, int H, int M) except +
+
+        void addPairwiseGaussian(float sx, float sy, LabelCompatibility*, KernelType, NormalizationType)
+        void addPairwiseBilateral(float sx, float sy, float sr, float sg, float sb, const unsigned char *rgbim, LabelCompatibility*, KernelType, NormalizationType)
+
+
+cdef class DenseCRF:
+    cdef c_DenseCRF *_this
+
+
+cdef class DenseCRF2D(DenseCRF):
+    cdef c_DenseCRF2D *_this2d
--- a/densecrf.pyx
+++ b/densecrf.pyx
+# distutils: language = c++
+# distutils: sources = densecrf/src/densecrf.cpp densecrf/src/unary.cpp densecrf/src/pairwise.cpp densecrf/src/permutohedral.cpp densecrf/src/optimization.cpp densecrf/src/objective.cpp densecrf/src/labelcompatibility.cpp densecrf/src/util.cpp densecrf/external/liblbfgs/lib/lbfgs.c
+# distutils: include_dirs = densecrf/include densecrf/external/liblbfgs/include
+
+from numbers import Number
+
+import eigen
+cimport eigen
+
+
+cdef LabelCompatibility* _labelcomp(compat):
+    if isinstance(compat, Number):
+        return new PottsCompatibility(compat)
+    elif memoryview(compat).ndim == 1:
+        return new DiagonalCompatibility(eigen.c_vectorXf(compat))
+    elif memoryview(compat).ndim == 2:
+        return new MatrixCompatibility(eigen.c_matrixXf(compat))
+    else:
+        raise ValueError("LabelCompatibility of dimension >2 not meaningful.")
+
+
+cdef class Unary:
+
+    # Because all of the APIs that take an object of this type will
+    # take ownership. Thus, we need to make sure not to delete this
+    # upon destruction.
+    cdef UnaryEnergy* move(self):
+        ptr = self.thisptr
+        self.thisptr = NULL
+        return ptr
+
+    # It might already be deleted by the library, actually.
+    # Yeah, pretty sure it is.
+    def __dealloc__(self):
+        del self.thisptr
+
+
+cdef class ConstUnary(Unary):
+    def __cinit__(self, float[:,::1] u not None):
+        self.thisptr = new ConstUnaryEnergy(eigen.c_matrixXf(u))
+
+
+cdef class LogisticUnary(Unary):
+    def __cinit__(self, float[:,::1] L not None, float[:,::1] f not None):
+        self.thisptr = new LogisticUnaryEnergy(eigen.c_matrixXf(L), eigen.c_matrixXf(f))
+
+
+cdef class DenseCRF:
+
+    def __cinit__(self, int nvar, int nlabels, *_, **__):
+        # We need to swallow extra-arguments because superclass cinit function
+        # will always be called with the same params as the subclass, automatically.
+
+        # We also only want to avoid creating an object if we're just being called
+        # from a subclass as part of the hierarchy.
+        if type(self) is DenseCRF:
+            self._this = new c_DenseCRF(nvar, nlabels)
+        else:
+            self._this = NULL
+
+    def __dealloc__(self):
+        # Because destructors are virtual, this is enough to delete any object
+        # of child classes too.
+        if self._this:
+            del self._this
+
+    def addPairwiseEnergy(self, float[:,::1] features not None, compat, KernelType kernel=DIAG_KERNEL, NormalizationType normalization=NORMALIZE_SYMMETRIC):
+        self._this.addPairwiseEnergy(eigen.c_matrixXf(features), _labelcomp(compat), kernel, normalization)
+
+    def setUnary(self, Unary u):
+        self._this.setUnaryEnergy(u.move())
+
+    def setUnaryEnergy(self, float[:,::1] u not None, float[:,::1] f = None):
+        if f is None:
+            self._this.setUnaryEnergy(eigen.c_matrixXf(u))
+        else:
+            self._this.setUnaryEnergy(eigen.c_matrixXf(u), eigen.c_matrixXf(f))
+
+    def inference(self, int niter):
+        return eigen.MatrixXf().wrap(self._this.inference(niter))
+
+    def startInference(self):
+        return eigen.MatrixXf().wrap(self._this.startInference()), eigen.MatrixXf(), eigen.MatrixXf()
+
+    def stepInference(self, MatrixXf Q, MatrixXf tmp1, MatrixXf tmp2):
+        self._this.stepInference(Q.m, tmp1.m, tmp2.m)
+
+    def klDivergence(self, MatrixXf Q):
+        return self._this.klDivergence(Q.m)
+
+
+cdef class DenseCRF2D(DenseCRF):
+
+    # The same comments as in the superclass' `__cinit__` apply here.
+    def __cinit__(self, int w, int h, int nlabels, *_, **__):
+        if type(self) is DenseCRF2D:
+            self._this = self._this2d = new c_DenseCRF2D(w, h, nlabels)
+
+    def addPairwiseGaussian(self, sxy, compat, KernelType kernel=DIAG_KERNEL, NormalizationType normalization=NORMALIZE_SYMMETRIC):
+        if isinstance(sxy, Number):
+            sxy = (sxy, sxy)
+
+        self._this2d.addPairwiseGaussian(sxy[0], sxy[1], _labelcomp(compat), kernel, normalization)
+
+    def addPairwiseBilateral(self, sxy, srgb, unsigned char[:,:,::1] rgbim not None, compat, KernelType kernel=DIAG_KERNEL, NormalizationType normalization=NORMALIZE_SYMMETRIC):
+        if isinstance(sxy, Number):
+            sxy = (sxy, sxy)
+
+        if isinstance(srgb, Number):
+            srgb = (srgb, srgb, srgb)
+
+        self._this2d.addPairwiseBilateral(
+            sxy[0], sxy[1], srgb[0], srgb[1], srgb[2], &rgbim[0,0,0], _labelcomp(compat), kernel, normalization
+        )
--- a/densecrf/README.txt
+++ b/densecrf/README.txt
+DenseCRF - Code
+=============
+http://graphics.stanford.edu/projects/drf/
+
+This software pertains to the research described in the ICML 2013 paper:
+Parameter Learning and Convergent Inference for Dense Random Fields, by
+Philipp Krähenbühl and Vladlen Koltun
+and the NIPS 2011 paper:
+Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials, by
+Philipp Krähenbühl and Vladlen Koltun
+
+If you're using this code in a publication, please cite our papers.
+
+This software is provided for research purposes, with absolutely no warranty
+or suggested support, and use of it most follow the BSD license agreement, at
+the top of each source file. *Please do not contact the authors for assistance
+with installing, understanding or running the code.* However if you think you
+have found an interesting bug, the authors would be grateful if you could pass
+on the information.
+
+Changes to the original code
+----------------------------
+The only major difference in this released version of the code is, that I directly
+compute the gradient of the permutohedral lattice, instead of the general Gauss
+Transform (3 line formula in p.6 in ICML 2013 paper). The gradient of the
+permutohedral lattice evaluated the exact gradient of the approximate filter.
+In higher dimensions (>3) the filter can be non continuous, which can complicate
+the optimization. The kernel gradient is also scaled lower than other parameters,
+which complicates the optimization. 
+
+
+How to compile the code
+-----------------------
+Dependencies:
+ * cmake  http://www.cmake.org/
+ * Eigen (included)
+ * liblbfgs (included)
+
+Linux, Mac OS X and Windows (cygwin):
+ mkdir build
+ cd build
+ cmake -D CMAKE_BUILD_TYPE=Release ..
+ make
+ cd ..
+
+Windows
+ You're probably better off just copying all files into a Visual Studio
+ project
+
+
+How to run the example
+----------------------
+An example on how to use the DenseCRF can be found in
+examples/dense_inference.cpp. The example loads an image and some annotations.
+It then uses a very simple classifier to compute a unary term based on those
+annotations. A dense CRF with both color dependent and color independent terms
+find the final accurate labeling.
+
+Linux, Mac OS X and Windows (cygwin):
+ build/examples/dense_inference input_image.ppm annotations.ppm output.ppm
+
+For example:
+ build/examples/dense_inference examples/im1.ppm examples/anno1.ppm output1.ppm
+
+
+An example on how to unse the learning code can be found in 
+examples/dense_learning.cpp. The example loads a color image and ground truth
+annotation. It then learn a CRF model with a logistic regression, a label comp
+and Gaussian kernel.
+
+Linux, Mac OS X and Windows (cygwin):
+ build/examples/dense_learning input_image.ppm annotations.ppm output.ppm
+
+For example:
+ build/examples/dense_learning examples/im1.ppm examples/anno1.ppm output1.ppm
+
+
+Please note that this implementation is slightly slower than the one used to
+in our NIPS 2011 paper. Mainly because I tried to keep the code clean and easy
+to understand.
+
--- a/densecrf/external/CMakeLists.txt
+++ b/densecrf/external/CMakeLists.txt
+include_directories( liblbfgs/include )
+add_library( lbfgs liblbfgs/lib/lbfgs.c )
--- a/densecrf/external/liblbfgs/AUTHORS
+++ b/densecrf/external/liblbfgs/AUTHORS
+Naoaki Okazaki <okazaki at chokkan org>
--- a/densecrf/external/liblbfgs/COPYING
+++ b/densecrf/external/liblbfgs/COPYING
+The MIT License
+
+Copyright (c) 1990 Jorge Nocedal
+Copyright (c) 2007-2010 Naoaki Okazaki
+
+Permission is hereby granted, free of charge, to any person obtaining a
+copy of this software and associated documentation files (the "Software"),
+to deal in the Software without restriction, including without limitation
+the rights to use, copy, modify, merge, publish, distribute, sublicense,
+and/or sell copies of the Software, and to permit persons to whom the
+Software is furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.
--- a/densecrf/external/liblbfgs/ChangeLog
+++ b/densecrf/external/liblbfgs/ChangeLog
+2010-xx-xx  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.10:
+	- Fixed compiling errors on Mac OS X; this patch was kindly submitted by Nic Schraudolph.
+	- Reduced compiling warnings on Mac OS X; this patch was kindly submitted by Tamas Nepusz.
+
+
+2010-01-29  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.9:
+	- Fixed a mistake in checking the validity of the parameters "ftol" and "wolfe"; this mistake was discovered by Kevin S. Van Horn.
+
+
+2009-07-13  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.8:
+	- Accepted the patch submitted by Takashi Imamichi; the backtracking method now has three criteria for choosing the step length.
+	- Updated the documentation to explain the above three criteria.
+
+
+2009-02-28  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.7:
+	- Improved OWL-QN routines for stability.
+	- Removed the support of OWL-QN method in MoreThuente algorithm
+	  because it accidentally fails in early stages of iterations for some
+	  objectives. Because of this change, the OW-LQN method must be used
+	  with the backtracking algorithm (LBFGS_LINESEARCH_BACKTRACKING), or
+	  the library returns LBFGSERR_INVALID_LINESEARCH.
+	- Renamed line search algorithms as follows:
+	    - LBFGS_LINESEARCH_BACKTRACKING: regular Wolfe condition.
+	    - LBFGS_LINESEARCH_BACKTRACKING_LOOSE: regular Wolfe condition.
+	    - LBFGS_LINESEARCH_BACKTRACKING_STRONG: strong Wolfe condition.
+	- Source code clean-up.
+
+
+2008-11-02  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.6:
+	- Improved line-search algorithm with strong Wolfe condition, which
+	  was contributed by Takashi Imamichi. This routine is now default for
+	  LBFGS_LINESEARCH_BACKTRACKING. The previous line search algorithm
+	  with regular Wolfe condition is still available as
+	  LBFGS_LINESEARCH_BACKTRACKING_LOOSE.
+	- Configurable stop index for L1-norm computation. A member variable
+	  lbfgs_parameter_t::orthantwise_end was added to specify the index
+	  number at which the library stops computing the L1 norm of the
+	  variables. This is useful to prevent some variables from being
+	  regularized by the OW-LQN method.
+	- A sample program written in C++ (sample/sample.cpp).
+
+
+2008-07-10  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.5:
+	- Configurable starting index for L1-norm computation. A member
+	  variable lbfgs_parameter_t::orthantwise_start was added to specify
+	  the index number from which the library computes the L1 norm of the
+	  variables.
+	- Fixed a zero-division error when the initial variables have already
+	  been a minimizer (reported by Takashi Imamichi). In this case, the
+	  library returns LBFGS_ALREADY_MINIMIZED status code.
+	- Defined LBFGS_SUCCESS status code as zero; removed unused constants,
+	  LBFGSFALSE and LBFGSTRUE.
+	- Fixed a compile error in an implicit down-cast.
+
+
+2008-04-25  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.4:
+	- Configurable line search algorithms. A member variable
+	  lbfgs_parameter_t::linesearch was added to choose either MoreThuente
+	  method (LBFGS_LINESEARCH_MORETHUENTE) or backtracking algorithm
+	  (LBFGS_LINESEARCH_BACKTRACKING).
+	- Fixed a bug: the previous version did not compute psuedo-gradients
+	  properly in the line search routines for OW-LQN. This bug might quit
+	  an iteration process too early when the OW-LQN routine was activated
+	  (0 < lbfgs_parameter_t::orthantwise_c).
+	- Configure script for POSIX environments.
+	- SSE/SSE2 optimizations with GCC.
+	- New functions lbfgs_malloc and lbfgs_free to use SSE/SSE2 routines
+	  transparently. It is uncessary to use these functions for libLBFGS
+	  built without SSE/SSE2 routines; you can still use any memory
+	  allocators if SSE/SSE2 routines are disabled in libLBFGS.
+
+
+2007-12-16  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.3:
+	- An API change. An argument was added to lbfgs() function to receive
+	  the final value of the objective function. This argument can be set
+	  to NULL if the final value is unnecessary.
+	- Fixed a null-pointer bug in the sample code (reported by Takashi
+	  Imamichi).
+	- Added build scripts for Microsoft Visual Studio 2005 and GCC.
+	- Added README file.
+	
+
+2007-12-13  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.2:
+	- Fixed a serious bug in orthant-wise L-BFGS. An important variable
+	  was used without initialization.
+	- Configurable L-BFGS parameters (number of limited memories, epsilon).
+	
+	
+2007-12-01  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.1:
+	- Implemented orthant-wise L-BFGS.
+	- Implemented lbfgs_parameter_init() function.
+	- Fixed several bugs.
+	- API documentation.
+	
+	
+2007-09-20  Naoaki Okazaki  <okazaki at chokkan org>
+
+	* libLBFGS 1.0
+	- Initial release.
+
--- a/densecrf/external/liblbfgs/INSTALL
+++ b/densecrf/external/liblbfgs/INSTALL
+Installation Instructions
+*************************
+
+Copyright (C) 1994, 1995, 1996, 1999, 2000, 2001, 2002, 2004 Free
+Software Foundation, Inc.
+
+This file is free documentation; the Free Software Foundation gives
+unlimited permission to copy, distribute and modify it.
+
+Basic Installation
+==================
+
+These are generic installation instructions.
+
+   The `configure' shell script attempts to guess correct values for
+various system-dependent variables used during compilation.  It uses
+those values to create a `Makefile' in each directory of the package.
+It may also create one or more `.h' files containing system-dependent
+definitions.  Finally, it creates a shell script `config.status' that
+you can run in the future to recreate the current configuration, and a
+file `config.log' containing compiler output (useful mainly for
+debugging `configure').
+
+   It can also use an optional file (typically called `config.cache'
+and enabled with `--cache-file=config.cache' or simply `-C') that saves
+the results of its tests to speed up reconfiguring.  (Caching is
+disabled by default to prevent problems with accidental use of stale
+cache files.)
+
+   If you need to do unusual things to compile the package, please try
+to figure out how `configure' could check whether to do them, and mail
+diffs or instructions to the address given in the `README' so they can
+be considered for the next release.  If you are using the cache, and at
+some point `config.cache' contains results you don't want to keep, you
+may remove or edit it.
+
+   The file `configure.ac' (or `configure.in') is used to create
+`configure' by a program called `autoconf'.  You only need
+`configure.ac' if you want to change it or regenerate `configure' using
+a newer version of `autoconf'.
+
+The simplest way to compile this package is:
+
+  1. `cd' to the directory containing the package's source code and type
+     `./configure' to configure the package for your system.  If you're
+     using `csh' on an old version of System V, you might need to type
+     `sh ./configure' instead to prevent `csh' from trying to execute
+     `configure' itself.
+
+     Running `configure' takes awhile.  While running, it prints some
+     messages telling which features it is checking for.
+
+  2. Type `make' to compile the package.
+
+  3. Optionally, type `make check' to run any self-tests that come with
+     the package.
+
+  4. Type `make install' to install the programs and any data files and
+     documentation.
+
+  5. You can remove the program binaries and object files from the
+     source code directory by typing `make clean'.  To also remove the
+     files that `configure' created (so you can compile the package for
+     a different kind of computer), type `make distclean'.  There is
+     also a `make maintainer-clean' target, but that is intended mainly
+     for the package's developers.  If you use it, you may have to get
+     all sorts of other programs in order to regenerate files that came
+     with the distribution.
+
+Compilers and Options
+=====================
+
+Some systems require unusual options for compilation or linking that the
+`configure' script does not know about.  Run `./configure --help' for
+details on some of the pertinent environment variables.
+
+   You can give `configure' initial values for configuration parameters
+by setting variables in the command line or in the environment.  Here
+is an example:
+
+     ./configure CC=c89 CFLAGS=-O2 LIBS=-lposix
+
+   *Note Defining Variables::, for more details.
+
+Compiling For Multiple Architectures
+====================================
+
+You can compile the package for more than one kind of computer at the
+same time, by placing the object files for each architecture in their
+own directory.  To do this, you must use a version of `make' that
+supports the `VPATH' variable, such as GNU `make'.  `cd' to the
+directory where you want the object files and executables to go and run
+the `configure' script.  `configure' automatically checks for the
+source code in the directory that `configure' is in and in `..'.
+
+   If you have to use a `make' that does not support the `VPATH'
+variable, you have to compile the package for one architecture at a
+time in the source code directory.  After you have installed the
+package for one architecture, use `make distclean' before reconfiguring
+for another architecture.
+
+Installation Names
+==================
+
+By default, `make install' will install the package's files in
+`/usr/local/bin', `/usr/local/man', etc.  You can specify an
+installation prefix other than `/usr/local' by giving `configure' the
+option `--prefix=PREFIX'.
+
+   You can specify separate installation prefixes for
+architecture-specific files and architecture-independent files.  If you
+give `configure' the option `--exec-prefix=PREFIX', the package will
+use PREFIX as the prefix for installing programs and libraries.
+Documentation and other data files will still use the regular prefix.
+
+   In addition, if you use an unusual directory layout you can give
+options like `--bindir=DIR' to specify different values for particular
+kinds of files.  Run `configure --help' for a list of the directories
+you can set and what kinds of files go in them.
+
+   If the package supports it, you can cause programs to be installed
+with an extra prefix or suffix on their names by giving `configure' the
+option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'.
+
+Optional Features
+=================
+
+Some packages pay attention to `--enable-FEATURE' options to
+`configure', where FEATURE indicates an optional part of the package.
+They may also pay attention to `--with-PACKAGE' options, where PACKAGE
+is something like `gnu-as' or `x' (for the X Window System).  The
+`README' should mention any `--enable-' and `--with-' options that the
+package recognizes.
+
+   For packages that use the X Window System, `configure' can usually
+find the X include and library files automatically, but if it doesn't,
+you can use the `configure' options `--x-includes=DIR' and
+`--x-libraries=DIR' to specify their locations.
+
+Specifying the System Type
+==========================
+
+There may be some features `configure' cannot figure out automatically,
+but needs to determine by the type of machine the package will run on.
+Usually, assuming the package is built to be run on the _same_
+architectures, `configure' can figure that out, but if it prints a
+message saying it cannot guess the machine type, give it the
+`--build=TYPE' option.  TYPE can either be a short name for the system
+type, such as `sun4', or a canonical name which has the form:
+
+     CPU-COMPANY-SYSTEM
+
+where SYSTEM can have one of these forms:
+
+     OS KERNEL-OS
+
+   See the file `config.sub' for the possible values of each field.  If
+`config.sub' isn't included in this package, then this package doesn't
+need to know the machine type.
+
+   If you are _building_ compiler tools for cross-compiling, you should
+use the `--target=TYPE' option to select the type of system they will
+produce code for.
+
+   If you want to _use_ a cross compiler, that generates code for a
+platform different from the build platform, you should specify the
+"host" platform (i.e., that on which the generated programs will
+eventually be run) with `--host=TYPE'.
+
+Sharing Defaults
+================
+
+If you want to set default values for `configure' scripts to share, you
+can create a site shell script called `config.site' that gives default
+values for variables like `CC', `cache_file', and `prefix'.
+`configure' looks for `PREFIX/share/config.site' if it exists, then
+`PREFIX/etc/config.site' if it exists.  Or, you can set the
+`CONFIG_SITE' environment variable to the location of the site script.
+A warning: not all `configure' scripts look for a site script.
+
+Defining Variables
+==================
+
+Variables not defined in a site shell script can be set in the
+environment passed to `configure'.  However, some packages may run
+configure again during the build, and the customized values of these
+variables may be lost.  In order to avoid this problem, you should set
+them in the `configure' command line, using `VAR=value'.  For example:
+
+     ./configure CC=/usr/local2/bin/gcc
+
+will cause the specified gcc to be used as the C compiler (unless it is
+overridden in the site shell script).
+
+`configure' Invocation
+======================
+
+`configure' recognizes the following options to control how it operates.
+
+`--help'
+`-h'
+     Print a summary of the options to `configure', and exit.
+
+`--version'
+`-V'
+     Print the version of Autoconf used to generate the `configure'
+     script, and exit.
+
+`--cache-file=FILE'
+     Enable the cache: use and save the results of the tests in FILE,
+     traditionally `config.cache'.  FILE defaults to `/dev/null' to
+     disable caching.
+
+`--config-cache'
+`-C'
+     Alias for `--cache-file=config.cache'.
+
+`--quiet'
+`--silent'
+`-q'
+     Do not print messages saying which checks are being made.  To
+     suppress all normal output, redirect it to `/dev/null' (any error
+     messages will still be shown).
+
+`--srcdir=DIR'
+     Look for the package's source code in directory DIR.  Usually
+     `configure' can determine that directory automatically.
+
+`configure' also accepts some other, not widely useful, options.  Run
+`configure --help' for more details.
+
--- a/densecrf/external/liblbfgs/Makefile.am
+++ b/densecrf/external/liblbfgs/Makefile.am
+# $Id$
+
+SUBDIRS = lib sample
+
+docdir = $(prefix)/share/doc/@PACKAGE@
+doc_DATA = README INSTALL COPYING AUTHORS ChangeLog NEWS
+
+EXTRA_DIST = \
+	autogen.sh \
+	lbfgs.sln
--- a/densecrf/external/liblbfgs/Makefile.in
+++ b/densecrf/external/liblbfgs/Makefile.in
--- a/densecrf/external/liblbfgs/NEWS
+++ b/densecrf/external/liblbfgs/NEWS
--- a/densecrf/external/liblbfgs/README
+++ b/densecrf/external/liblbfgs/README
+
+           libLBFGS: C library of limited-memory BFGS (L-BFGS)
+
+                                       Copyright (c) 1990, Jorge Nocedal
+                                 Copyright (c) 2007-2010, Naoaki Okazaki
+
+=========================================================================
+1. Introduction
+=========================================================================
+libLBFGS is a C port of the implementation of Limited-memory
+Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method written by Jorge Nocedal.
+The original FORTRAN source code is available at:
+http://www.ece.northwestern.edu/~nocedal/lbfgs.html
+
+The L-BFGS method solves the unconstrainted minimization problem:
+    minimize F(x), x = (x1, x2, ..., xN),
+only if the objective function F(x) and its gradient G(x) are computable.
+
+Refer to the libLBFGS web site for more information.
+http://www.chokkan.org/software/liblbfgs/
+
+
+
+=========================================================================
+2. How to build
+=========================================================================
+[Microsoft Visual Studio 2008]
+Open the solution file "lbfgs.sln" and build it.
+
+[GCC]
+$ ./configure
+$ make
+$ make install  # To install libLBFGS library and header.
+
+
+
+=========================================================================
+3. Note on SSE/SSE2 optimization
+=========================================================================
+This library has SSE/SSE2 optimization routines for vector arithmetic
+operations on Intel/AMD processors. The SSE2 routine is for 64 bit double
+values, and the SSE routine is for 32 bit float values. Since the default
+parameters in libLBFGS are tuned for double precision values, it may need
+to modify these parameters to use the SSE optimization routines.
+
+To use the SSE2 optimization routine, specify --enable-sse2 option to the
+configure script.
+
+$ ./configure --enable-sse2
+
+To build libLBFGS with SSE2 optimization enabled on Microsoft Visual
+Studio 2005, define USE_SSE and __SSE2__ symbols.
+
+Make sure to run libLBFGS on processors where SSE2 instrunctions are
+available. The library does not check the existence of SSE2 instructions.
+
+To package maintainers,
+
+Please do not enable SSE/SSE2 optimization routine. The library built
+with SSE/SSE2 optimization will crash without any notice when necessary
+SSE/SSE2 instructions are unavailable on CPUs.
+
+
+
+=========================================================================
+4. License
+=========================================================================
+libLBFGS is distributed under the term of the MIT license.
+Please refer to COPYING file in the distribution.
+
+$Id$
--- a/densecrf/external/liblbfgs/aclocal.m4
+++ b/densecrf/external/liblbfgs/aclocal.m4
--- a/densecrf/external/liblbfgs/autogen.sh
+++ b/densecrf/external/liblbfgs/autogen.sh
+#!/bin/sh
+# $Id$
+
+if [ "$1" = "--force" ];
+then
+    FORCE=--force
+    NOFORCE=
+    FORCE_MISSING=--force-missing
+else
+    FORCE=
+    NOFORCE=--no-force
+    FORCE_MISSING=
+fi
+
+libtoolize --copy $FORCE 2>&1 | sed '/^You should/d' || {
+    echo "libtoolize failed!"
+    exit 1
+}
+
+aclocal $FORCE || {
+    echo "aclocal failed!"
+    exit 1
+}
+
+autoheader $FORCE || {
+    echo "autoheader failed!"
+    exit 1
+}
+
+automake -a -c $NOFORCE || {
+    echo "automake failed!"
+    exit 1
+}
+
+autoconf $FORCE || {
+    echo "autoconf failed!"
+    exit 1
+}
--- a/densecrf/external/liblbfgs/config.guess
+++ b/densecrf/external/liblbfgs/config.guess
--- a/densecrf/external/liblbfgs/config.h.in
+++ b/densecrf/external/liblbfgs/config.h.in
+/* config.h.in.  Generated from configure.in by autoheader.  */
+
+/* Define to 1 if you have the <dlfcn.h> header file. */
+#undef HAVE_DLFCN_H
+
+/* Define to 1 if you have the <emmintrin.h> header file. */
+#undef HAVE_EMMINTRIN_H
+
+/* Define to 1 if you have the <inttypes.h> header file. */
+#undef HAVE_INTTYPES_H
+
+/* Define to 1 if you have the `m' library (-lm). */
+#undef HAVE_LIBM
+
+/* Define to 1 if you have the <memory.h> header file. */
+#undef HAVE_MEMORY_H
+
+/* Define to 1 if you have the <stdint.h> header file. */
+#undef HAVE_STDINT_H
+
+/* Define to 1 if you have the <stdlib.h> header file. */
+#undef HAVE_STDLIB_H
+
+/* Define to 1 if you have the <strings.h> header file. */
+#undef HAVE_STRINGS_H
+
+/* Define to 1 if you have the <string.h> header file. */
+#undef HAVE_STRING_H
+
+/* Define to 1 if you have the <sys/stat.h> header file. */
+#undef HAVE_SYS_STAT_H
+
+/* Define to 1 if you have the <sys/types.h> header file. */
+#undef HAVE_SYS_TYPES_H
+
+/* Define to 1 if you have the <unistd.h> header file. */
+#undef HAVE_UNISTD_H
+
+/* Define to 1 if you have the <xmmintrin.h> header file. */
+#undef HAVE_XMMINTRIN_H
+
+/* Name of package */
+#undef PACKAGE
+
+/* Define to the address where bug reports for this package should be sent. */
+#undef PACKAGE_BUGREPORT
+
+/* Define to the full name of this package. */
+#undef PACKAGE_NAME
+
+/* Define to the full name and version of this package. */
+#undef PACKAGE_STRING
+
+/* Define to the one symbol short name of this package. */
+#undef PACKAGE_TARNAME
+
+/* Define to the version of this package. */
+#undef PACKAGE_VERSION
+
+/* Define to 1 if you have the ANSI C header files. */
+#undef STDC_HEADERS
+
+/* Version number of package */
+#undef VERSION
--- a/densecrf/external/liblbfgs/config.sub
+++ b/densecrf/external/liblbfgs/config.sub
--- a/densecrf/external/liblbfgs/configure
+++ b/densecrf/external/liblbfgs/configure
--- a/densecrf/external/liblbfgs/configure.in
+++ b/densecrf/external/liblbfgs/configure.in
+dnl $Id$
+dnl
+dnl
+dnl Exported and configured variables:
+dnl CFLAGS
+dnl LDFLAGS
+dnl INCLUDES
+
+
+dnl ------------------------------------------------------------------
+dnl Initialization for autoconf
+dnl ------------------------------------------------------------------
+AC_PREREQ(2.59)
+AC_INIT
+AC_CONFIG_SRCDIR([lib/lbfgs.c])
+
+
+dnl ------------------------------------------------------------------
+dnl Initialization for automake
+dnl ------------------------------------------------------------------
+AM_INIT_AUTOMAKE(liblbfgs, 1.10)
+AC_CONFIG_HEADERS(config.h)
+AM_MAINTAINER_MODE
+
+
+dnl ------------------------------------------------------------------
+dnl Checks for program
+dnl ------------------------------------------------------------------
+AC_PROG_LIBTOOL
+AC_PROG_INSTALL
+AC_PROG_LN_S
+AC_PROG_MAKE_SET
+
+
+dnl ------------------------------------------------------------------
+dnl Initialization for variables
+dnl ------------------------------------------------------------------
+CFLAGS="${ac_save_CFLAGS} -Wall"
+LDFLAGS="${ac_save_LDFLAGS}"
+INCLUDES="-I\$(top_srcdir) -I\$(top_srcdir)/include"
+
+
+
+dnl ------------------------------------------------------------------
+dnl Checks for header files.
+dnl ------------------------------------------------------------------
+AC_HEADER_STDC
+AC_CHECK_HEADERS(xmmintrin.h emmintrin.h)
+
+
+
+dnl ------------------------------------------------------------------
+dnl Checks for debugging mode
+dnl ------------------------------------------------------------------
+AC_ARG_ENABLE(
+    debug,
+    [AS_HELP_STRING(
+        [--enable-debug],
+        [build for debugging]
+        )],
+    [CFLAGS="-DDEBUG -O -g ${CFLAGS}"],
+    [CFLAGS="-O3 -ffast-math ${CFLAGS}"]
+)
+
+dnl ------------------------------------------------------------------
+dnl Checks for profiling mode
+dnl ------------------------------------------------------------------
+AC_ARG_ENABLE(
+    profile,
+    [AS_HELP_STRING(
+        [--enable-profile],
+        [build for profiling]
+	)],
+    [CFLAGS="-DPROFILE -pg ${CFLAGS}"]
+)
+
+dnl ------------------------------------------------------------------
+dnl Checks for SSE2 build
+dnl ------------------------------------------------------------------
+AC_ARG_ENABLE(
+    sse2,
+    [AS_HELP_STRING(
+        [--enable-sse2],
+        [enable SSE2 optimization routines]
+        )],
+    [CFLAGS="-msse2 -DUSE_SSE ${CFLAGS}"]
+)
+
+dnl ------------------------------------------------------------------
+dnl Checks for library functions.
+dnl ------------------------------------------------------------------
+AC_CHECK_LIB(m, fabs)
+
+
+
+dnl ------------------------------------------------------------------
+dnl Export variables
+dnl ------------------------------------------------------------------
+AC_SUBST(CFLAGS)
+AC_SUBST(LDFLAGS)
+AC_SUBST(INCLUDES)
+
+dnl ------------------------------------------------------------------
+dnl Output the configure results.
+dnl ------------------------------------------------------------------
+AC_CONFIG_FILES(Makefile lib/Makefile sample/Makefile)
+AC_OUTPUT