Darwin  1.10(beta)
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Groups Pages
Class List
Here are the classes, structs, unions and interfaces with brief descriptions:
[detail level 12]
oNdrwnFactorGraphUtilsUtility functions for factor graphs
oNdrwnMatlabUtilsUtility functions for converting between Matlab and Darwin datatypes
oNdrwnNNGraphMovesTemplated drwnNNGraph search moves
oNdrwnNNGraphThreadedMovesTemplated drwnNNGraph search moves (multi-threaded)
oNdrwnNNGraphVisVisualization routines for a drwnNNGraph
oNdrwnPatchMatchVisVisualization routines
oCdrwnADLPInferenceImplements the alternating direction method algorithm described in "An Alternating Direction Method for Dual MAP LP Relaxation," Ofer Meshi and Amir Globerson, ECML, 2011
oCdrwnAlphaBetaSwapInferenceImplements alpha-beta swap inference using graph-cuts (see Boykov et al, 2001). Factor graphs must be pairwise
oCdrwnAlphaExpansionInferenceImplements alpha-expansion inference using graph-cuts (see Boykov et al, 2001). Factor graphs must be pairwise
oCdrwnAsyncMaxProdInferenceImplements asynchronous max-product (min-sum) inference
oCdrwnAsyncSumProdInferenceImplements asynchronous sum-product inference
oCdrwnBiasFeatureMapAugments input feature vector with 1 (i.e., to allow for a bias weight)
oCdrwnBiasJointFeatureMapSame as drwnIdentityJointFeatureMap but adds a bias term for each class i.e., $\phi(x, y) = \left(\delta\!\left\{y = 0\right\} (x^T, 1), \ldots, \delta\!\left\{y = K - 2\right\} (x^T, 1)\right) \in \mathbb{R}^{(K - 1)(n + 1)}$
oCdrwnBitArrayImplements an efficient packed array of bits
oCdrwnBKMaxFlowImplementation of Boykov and Kolmogorov's maxflow algorithm
oCdrwnBoostedClassifierImplements a mult-class boosted decision-tree classifier. See Zhu et al., Multi-class AdaBoost, 2006
oCdrwnClassificationResultsEncapsulates summary of classifier output from which various curves can be generated (e.g., precision-recall curves)
oCdrwnClassifierImplements the interface for a generic machine learning classifier
oCdrwnCloneableInterface for cloning object (i.e., virtual copy constructor)
oCdrwnCodeProfilerStatic class for providing profile information on functions
oCdrwnColourHistogramSpecialized histogram for quantized 3-channel colour values (e.g., RGB)
oCdrwnCompositeClassifierImplements a multi-class classifier by combining binary classifiers
oCdrwnCompressionBufferUtility class for compressing data using the zlib library
oCdrwnConditionalGaussianUtility class for generating conditonal gaussian distribution
oCdrwnCondSuffStatsImplements a class for accumulating conditional first- and second-order sufficient statistics
oCdrwnConfigurableModuleInterface for a configurable module
oCdrwnConfigurationManagerConfiguration manager
oCdrwnConfusionMatrixUtility class for computing and printing confusion matrices
oCdrwnCrossValidatorUtility class for cross-validating classifier meta-parameters by brute-force testing of all combinations of some given settings
oCdrwnDatasetImplements a cacheable dataset containing feature vectors, labels and optional weights
oCdrwnDecisionTreeImplements a (binary-split) decision tree classifier of arbitrary depth
oCdrwnDeformationCostStructure for holding dx, dy, dx^2 and dy^2 deformation costs
oCdrwnDisjointSetsImplements a forest of disjoint sets abstract data type
oCdrwnDualDecompositionInferenceImplements dual decomposition MAP inference (see Komodakis and Paragios, CVPR 2009 and works cited therein). Each factor is treated as a separate slave
oCdrwnEdmondsKarpMaxFlowImplementation of Edmonds-Karp maxflow/min-cut algorithm
oCdrwnFactorGeneric interface for a factor. Currently only inherited by drwnTableFactor
oCdrwnFactorAdditionOpAdd two or more factors together
oCdrwnFactorAtomicOpExecutes an atomic operation by executing a sequence of factor operations
oCdrwnFactorBinaryOpBase class for implementing binary factor operations
oCdrwnFactorCopyOpCopy one factor onto another
oCdrwnFactorDivideOpDivide one factor by another
oCdrwnFactorExpAndNormalizeOpExponentiate and normalize all the entries in a factor to sum to one
oCdrwnFactorGraphContainer and utility functions for factor graphs
oCdrwnFactorLogNormalizeOpShift all the entries in a factor so that the maximum is zero
oCdrwnFactorMarginalizeOpMarginalize out one or more variables in a factor
oCdrwnFactorMaximizeOpMaximize over one or more variables in a factor
oCdrwnFactorMinimizeOpMinimize over one or more variables in a factor
oCdrwnFactorMinusEqualsOpSubtract one (weighted) factor from another inline
oCdrwnFactorNAryOpBase class for implementing n-ary factor operations
oCdrwnFactorNormalizeOpNormalize all the entries in a factor to sum to one. Assumes non-negative entries
oCdrwnFactorOperationBase class for implementing various operations on table factors. The derived classes store mappings between factor entries making them very fast for iterative algorithms
oCdrwnFactorPlusEqualsOpAdd one (weighted) factor to another inline
oCdrwnFactorProductOpMultiply two or more factors together
oCdrwnFactorReduceOpReduce factor by oberving the value of one or more variables
oCdrwnFactorSubtractOpSubtract one factor from another
oCdrwnFactorTimesEqualsOpMultiply one factor by another inline
oCdrwnFactorUnaryOpBase class for implementing unary factor operations
oCdrwnFactorWeightedSumOpAdd a weighted combination of factors
oCdrwnFactoryTemplated factory for creating or cloning objects for a particular base class
oCdrwnFactoryAutoRegisterHelper class for registering classes with a drwnFactory
oCdrwnFactoryTraitsSome classes may provide default factory registration (e.g., built-in classes such as drwnClassifier and drwnFeatureTransform)
oCdrwnFactoryTraits< drwnClassifier >Implements factory for classes derived from drwnClassifier with automatic registration of built-in classes
oCdrwnFactoryTraits< drwnFeatureTransform >Implements factory for classes derived from drwnFeatureTransform with automatic registration of built-in classes
oCdrwnFeatureMapDefines the interface for a feature mapping $\phi : \mathbb{R}^n \rightarrow \mathbb{R}^m$
oCdrwnFeatureTransformImplements the interface for a generic feature transforms possibly with learned parameters, e.g., PCA (unsupervised) or LDA (supervised)
oCdrwnFeatureWhitenerWhitens (zero mean, unit variance) feature vector (see also drwnPCA)
oCdrwnFilterBankResponseHolds the results of running an image through a bank of filters and allows for computation of features over rectangular regions
oCdrwnFisherLDAFisher's linear discriminant analysis (LDA)
oCdrwnGaussianImplements a multi-variate gaussian distribution
oCdrwnGaussianMixtureImplements a multi-variant Gaussian mixture model
oCdrwnGEMPLPInferenceImplements the generalized LP-based message passing algorithm of Globerson and Jaakkola, NIPS 2007
oCdrwnGrabCutInstanceImplements the grabCut algorithm of Rother et al., SIGGRAPH 2004 for figure/ground segmentation
oCdrwnHistogramImplements a simple templated histogram class
oCdrwnHOGFeaturesEncapsulates histogram-of-gradient (HOG) feature computation
oCdrwnICMInferenceImplements iterated conditional modes (ICM) MAP inference. This method was first proposed in Besag, Royal Stats Society, 1986
oCdrwnIdentityFeatureMapCopies input feature space to output feature space
oCdrwnIdentityJointFeatureMapIncludes a copy of each feature from the input space for each class other than the last, i.e., $\phi(x, y) = \left(\delta\!\left\{y = 0\right\} x^T, \ldots, \delta\!\left\{y = K - 2\right\}x^T\right) \in \mathbb{R}^{(K - 1) n}$. This is the standard feature mapping for multi-class logistic models
oCdrwnImageCacheCaches images in memory up to a maximum number of images or memory limit
oCdrwnImageInPainterPerforms exemplar-based image inpainting
oCdrwnImagePyramidCacheCaches image pyramids in main memory up to a maximum number of images or memory limit
oCdrwnIndexQueueProvides a queue datastructure on a fixed number of indexes. At most one copy of each index can appear in the queue (a second enqueue is ignored). Membership of the queue can be queried
oCdrwnInferenceInterface for various (marginal) inference algorithms
oCdrwnJointFeatureMapDefines the interface for a joint feature mapping $\phi : \mathbb{R} \times \mathbb{Z} \rightarrow \mathbb{R}^m$
oCdrwnJunctionTreeInferenceImplements the junction tree algorithm for exact inference on a factor graph using drwnAsyncMaxProdInference for the actual message passing
oCdrwnKMeansImplements k-means clustering. Outputs the squared-distance to each of the cluster centroids. The nearest cluster can be found by passing the output to the drwn::argmin function. Supports weighted training examples
oCdrwnLBPFilterBankImplements filter bank for encoding local binary patterns
oCdrwnLinearRegressorBaseCommon functionality for drwnLinearRegressor
oCdrwnLinearTransformImplements a linear feature transform with externally settable parameters
oCdrwnLoggerMessage and error logging. This class is not thread-safe in the interest of not having to flush the log on every message
oCdrwnLPSolverSolves equality constrained linear programs with positivity constraints via the log-barrier method
oCdrwnMAPInferenceInterface for various MAP inference (energy minimization) algorithms
oCdrwnMAPInferenceFactoryFactory for creating drwnMAPInference objects
oCdrwnMaskedPatchMatchImplements the basic PatchMatch algorithm of Barnes et al., SIGGRAPH 2009 on masked images
oCdrwnMaxFlowInterface for maxflow/min-cut algorithms (for minimizing submodular quadratic pseudo-Boolean functions)
oCdrwnMaxProdInferenceImplements max-product inference
oCdrwnMessagePassingInferenceImplements generic message-passing algorithms on factor graphs. See derived classes for specific algorithms
oCdrwnMessagePassingMAPInferenceImplements generic message-passing algorithms on factor graphs. See derived classes for specific algorithms
oCdrwnMouseStateMouse state and mouse callback for populating the mouse state. Used by the drwnWaitMouse function
oCdrwnMultiClassLogisticBaseCommon functionality for drwnMultiClassLogistic
oCdrwnMultiSegConfigManages configuration settings for multiple image segmentation
oCdrwnMultiSegRegionDefinitionsProvides a mechanism for mapping region IDs to colours and class names. Can be initialized from an XML configuration file or programmatically for a number of standard datasets
oCdrwnNNGraphClass for maintaining a nearest neighbour graph over superpixel images. Search moves are implemented by templated functions in the drwnNNGraphMoves namespace
oCdrwnNNGraphDefaultMetricImplements the scoring functions needed by the search moves. Required member functions are isMatchable(), isFinite() and score()
oCdrwnNNGraphEdgeEncapsulates an outgoing edge in a drwnNNGraph
oCdrwnNNGraphImageHolds nodes (superpixels) for a single image
oCdrwnNNGraphImageDataHolds image, segments and other housekeeping information for an image
oCdrwnNNGraphLearnerLearn the distance metric base class with full set of constraints (i.e., loss function over all targets and imposters)
oCdrwnNNGraphLLearnerLearn the distance metric M = LL^T as L^T
oCdrwnNNGraphNodeEncapsulates a superpixel node in a drwnNNGraph
oCdrwnNNGraphNodeAnnotationTemplated utility class for holding annotations for every node in a graph. See learning code for example use
oCdrwnNNGraphSparseLearnerLearn the distance metric base class with sparse set of constraints (i.e., loss function over further target and nearest imposter only)
oCdrwnObjectEncapsulates a 2D object in an image for object detection
oCdrwnObjectListList of objects for the same image (see drwnObject)
oCdrwnObjectSequenceSequence of images, each with a list of objects (see drwnObjectList)
oCdrwnOptimizerInterface for solving large-scale unconstrained optimization problems using L-BFGS
oCdrwnOrderedMapProvides a datastructure for that can be indexed by a KeyType (usually a string) or unsigned integer, i.e., the index
oCdrwnPartA part is defined as a template over a number of channels (possibly one), an offset from the object centroid, and a deformation cost. The part is scored as w^T F(x) where w are the template weights, and F(x) is the feature vector at location x in feature space. Note for edge templates this can be in pixels, but generally will be in some multiple of pixels. It is up to the inference model to handle conversion from feature space to pixel space
oCdrwnPartialAssignmentDefines an assignment to a subset of the variables
oCdrwnPartsAssignmentClass for holding as assignment to part locations and occlusion variables
oCdrwnPartsInferenceHelper class for running inference in a (constellation) parts-based model. Supports linear and quadratic distance transforms for deformation costs. Computes argmax_{x,c,z} m_i(x_i, z_i) + d_i(x_i, c) + p(c) where m_i(x, 0) = matchingCost_i(x) + priorCost_i(x) m_i(x, 1) = occlusionCost + priorCost_i(x) d_i(x, c) = [dx, dy]^T fabs(x - o - c) + [dx2, dy2]^T (x - o - c).^2
oCdrwnPartsModelInterface for implementing a part-based constellation model (i.e., pictorial structures model) for object detection
oCdrwnPartsModelMixtureMixture of parts models (T must have a parts model interface). Inference returns the best scoring model and its parts locations
oCdrwnPatchMatchEdgeRepresents an edge in the drwnPatchMatchGraph
oCdrwnPatchMatchGraphEach image maintains a W-by-H-by-K array of match records referencing the (approximate) best K matches to other images. Image filename rather than the images themselves are stored. Duplicate filenames (images) are not allowed
oCdrwnPatchMatchGraphLearnerLearns a PatchMatchGraph by iteratively performing search moves over the space of matches
oCdrwnPatchMatchGraphRepaintClass for repainting an image from matches within the PatchMatchGraph
oCdrwnPatchMatchImagePyramidRecord of patch matches for mutliple levels each image
oCdrwnPatchMatchImageRecordRecords matches for one level in an image pyramid
oCdrwnPatchMatchNodeRepresents a node in the drwnPatchMatchGraph
oCdrwnPCAPrincipal component analysis feature transformation
oCdrwnPersistentBlockPersistent storage block used internally by drwnPersistentStorage
oCdrwnPersistentRecordInterface class for drwnPersistentStorage
oCdrwnPersistentStorageProvides indexed storage for multiple records using two files (a binary data file and a text index file)
oCdrwnPersistentStorageBufferProvides buffered storage (delayed write-back) of objects with a drwnPersistentRecord interface
oCdrwnPersistentVectorRecordTemplated class for storing vector records
oCdrwnPersistentVectorVectorRecordTemplated class for storing vector-of-vector records
oCdrwnPixelNeighbourContrastsConvenience class for holding pixel contrast weights
oCdrwnPixelSegCRFInferenceAlpha-expansion inference for a pixel-level CRF model with unary, contrast-dependent pairwise, and custom higher-order terms
oCdrwnPixelSegModelImplements a pixel-level CRF model for multi-class image segmentation (pixel labeling)
oCdrwnPRCurvePrecision-recall curve
oCdrwnPropertiesProvides an abstract interface for dynamic properties
oCdrwnQPSolverQuadratic program solver
oCdrwnQuadraticFeatureMapAugments input feature vector with square of each feature (normalized so that if input is zero mean and unit variance so will output) as well as cross-terms and constant one (i.e., to allow for a bias weight)
oCdrwnQuadraticJointFeatureMapSame as drwnSquareJointFeatureMap but adds cross-terms
oCdrwnRandomForestImplements a Random forest ensemble of decision trees classifier. See L. Breiman, "Random Forests", Machine Learning, 2001
oCdrwnRegressionImplements the interface for a generic machine learning regression, e.g. see drwnLinearRegressor
oCdrwnRobustPottsCRFInferenceAnd where 0.0 < < 0.5
oCdrwnSegImageCompositePixelFeaturesClass for generating composite per-pixel feature vectors
oCdrwnSegImageFilePixelFeaturesPre-processed per-pixel features stored in files
oCdrwnSegImageInstanceEncapsulates a single instance of an image for multi-class pixel labeling problems (i.e., image segmentation)
oCdrwnSegImagePixelFeaturesInterface for generating per-pixel features for a drwnSegImageInstance object
oCdrwnSegImageRegionFeaturesInterface for generating per-region (or per-superpixel) features for a drwnSegImageInstance object. The superpixel data member of the drwnSegImageInstance object must be populated
oCdrwnSegImageStdPixelFeaturesStandard per-pixel filterbank features with option to read auxiliary features from a file
oCdrwnSegImageStdRegionFeaturesStandard per-region filterbank features computes mean and standard deviation of drwnTextonFilterBank responses over each region
oCdrwnSmartPointerImplements a shared pointer interface to avoid the need to deep copy constant (shared) objects
oCdrwnSmartPointerCmpLessThanComparison operator for objects held in a drwnSmartPointer
oCdrwnSontag08InferenceImplements the incremental tightening of the LP MAP inference algorithm from Sontag et al., UAI 2008
oCdrwnSparseLPSolverSolves linear programs with sparse equality constraints
oCdrwnSparseVecQuick-and-dirty sparse vector class as a plugin replacement for std::vector
oCdrwnSquareFeatureMapAugments input feature vector with square of each feature (normalized so that if input is zero mean and unit variance so will output) and 1 (i.e., to allow for a bias weight)
oCdrwnSquareJointFeatureMapSame as drwnIdentityJointFeatureMap but adds a square term for each feature i.e., $\phi(x, y) = \left(\delta\!\left\{y = 0\right\} (x^T, \frac{1}{\sqrt{2}} diag(xx^T - I), 1), \ldots, \delta\!\left\{y = K - 2\right\} (x^T, \frac{1}{\sqrt{2}} diag(xx^T - I), 1)\right) \in \mathbb{R}^{(K - 1)(2n + 1)}$
oCdrwnStdObjIfaceStandard Darwin object interface (cloneable and writeable)
oCdrwnSuffStatsImplements a class for accumulating first- and second-order sufficient statistics (moments)
oCdrwnSumProdInferenceImplements sum-product inference
oCdrwnSuperpixelContainerHolds multiple oversegmentations for a given image
oCdrwnSupervisedTransformImplements interface for supervised feature transforms (i.e., with class labels)
oCdrwnTableFactorFactor which stores the value of each assignment explicitly in table form
oCdrwnTableFactorMappingCreates a mapping between entries in two tables
oCdrwnTableFactorStorageShared memory for table factors
oCdrwnTemplateMatcherUtility class for computing multiple template matches
oCdrwnTextonFilterBankImplements a 17-dimensional filter bank
oCdrwnTFeatureMapTransformHelper feature transformation based on a drwnFeatureMap
oCdrwnThreadJobInterface for a thread job functor
oCdrwnThreadPoolImplements a pool of threads for running concurrent jobs
oCdrwnTLinearRegressorImplements linear regression optimization templated on a drwnFeatureMap
oCdrwnTMultiClassLogisticImplements a multi-class logistic classifier templated on a drwnJointFeatureMap
oCdrwnTripletBasic datatype for holding three objects of arbitrary type. Similar to the STL pair<> class
oCdrwnTRWSInferenceImplements the sequential tree-reweighted message passing (TRW-S) algorithm described in "Convergent Tree-Reweighted Message Passing for Energy Minimization," Kolmogorov, IEEE PAMI, 2006
oCdrwnTypeableInterface for an object that returns its own type as a string
oCdrwnUnsupervisedTransformImplements interface for unsupervised feature transforms (i.e, without class labels)
oCdrwnVarUniverseData structure for definining the random variables (name and cardinality) for a given problem instance
oCdrwnWeightedPixelEdgeWeighted undirected arc between pixels in an image
oCdrwnWeightedRobustPottsCRFInferenceAnd where 0.0 < < 0.5, W_c = w_i
oCdrwnWriteableInterface for objects that can serialize and de-serialize themselves