====== Catalog of methods in AI/vision/robotics ======



This is a classification of techniques and algorithms, in order to give a broad view of solutions available to deal with classical problems.



Other pages contain some more details about, which can be seen as memos with references to find more information (with the origin paper when possible).



===== Learning =====
==== Supervised learning ====
=== Classification ===

  * __MLP (Multi Layers Perceptron)__ - //PMC (Perceptron multicouches)//
    * gradient backpropagation - //rétropropagation du gradient//
      * stochastic
      * with inertia
      * simulated annealing - //recuit simulé//
    * newton (second order)
  * __RBFNN (Radial Basis Functions Neural Networks)__
    * k-means then gradient descent
    * incremental addition of neurons then exact method
  * __SVM (Support Vectors Machine)__
  * __Decision tree__ - //arbre de décision//
    * ID3 (based on entropy)
  * __k-nearest neighbors__ - //k plus proches voisins//
  * __Boosting__
    * Boosting by majority
    * AdaBoost (ADAptive BOOSTing)
      * Discrete AdaBoost
      * Real AdaBoost
      * Gentle AdaBoost
      * LogitBoost
      * Probabilistic AdaBoost
      * FloatBoost
      * AdaBoost.Reg
      * Multiclass AdaBoost.M1
      * Multiclass AdaBoost.M2
      * Multilabel AdaBoost.MR
      * Multilabel AdaBoost.MH
      * Multiclass AdaBoost.MO
      * Multiclass AdaBoost.OC
      * Multiclass AdaBoost.ECC
      * GrPloss
      * BoostMA
      * AdaBoost.M1W
      * SAMME (Stagewise Additive Modeling using a Multi-class Exponential loss function)
      * GAMBLE (Gentle Adaptive Multiclass Boosting Learning)
    * UBoost
    * LPBoost (Linear Programming BOOSTing)
    * TotalBoost (TOTALly corrective BOOSTing)
    * RotBoost
    * alphaBoost
    * MILBoost (Multiple Instance Learning BOOSting)
    * CGBoost (Conjugate Gradient BOOSTing)
    * Bootstrap Aggregating
  * __Cascades of detectors__ [[classification#cascades_of_detectors|[+] ]]
  * __Trees of detectors__



=== Regression ===

  * __MLP (Multi Layers Perceptron)__
  * __RBFNN (Radial Basis Functions Neural Network)__
  * __SVR (Support Vectors Regressor)__

=== Pattern recognition ===

  * __Viola-Jones Detector__ [[pattern-recognition#viola-jones_detector|[+] ]]
    * with Extended Set of Haar features
    * Stumps or CART trees
    * Rotation Invariant
    * **Multiview**
      * Parallel Cascades
      * Pyramid Cascade
      * Tree Cascade
      * Vector Boosting

==== Unsupervised learning ====
=== Vector quantization / Clustering ===

  * Sequential leader
  * k-means - //k-moyennes//
  * GNG (Growing Neural Gas)
  * Auto-organizing maps (Kohonen) - //cartes auto-organisatrices de Kohonen//


==== Reinforcement learning ====

  * __MDP (Markov Decision Processes)__
    * Q-learning
    * Value iteration
    * Policy iteration

===== Planification =====
==== Symbolic ====
=== State space search ===

  * A*
    * WA*
    * IDA*
  * Dijkstra

=== Logics ===

  * __GraphPlan__
    * Stan
    * IPP
    * SGP
  * __SATplan (SATisfiability PLANning)__

==== Others ====

  * __Genetic algorithms__ - //algorithmes génétiques//
  * __Ant colonies__ - //colonies de fourmis//

==== Specific ====
=== Path planning ===

  * __Configurations space__
  * __Potential fields__

===== Perception =====
==== Vision ====
=== Color Quantization ===

  * RGB cone
  * YUV polygon
  * HSV rectangle
  * Lab

=== Image segmentation ===

  * Floodfill - //inondation//
  * Watershed - //lignes de partage des eaux//

=== Filters ===

  * **Anti-noise (smoothing)**
    * Median Filter - //filtre médian//
    * Vector Median Filter
    * Kuwahara filter
    * Peer Group Filtering
    * Anisotropic Filtering
  * **Gradient** [[vision#gradient|[+] ]]
    * Prewitt [[vision#prewitt|[+] ]]
    * Roberts [[vision#roberts|[+] ]]
    * Sobel [[vision#sobel|[+] ]]
    * Laplace [[vision#laplace|[+] ]]
    * Scharr [[vision#scharr|[+] ]]
  * **Morphological**
    * dilation - //dilatation//
    * erosion - //érosion//
    * opening - //ouverture//
    * closing - //fermeture//

=== Edge detection ===

  * Canny detector
    * Canny-Deriche

=== Pattern recognition ===

  * Mean-Square Regression - //Régression aux moindres carrés//
  * __Hough Transforms__ [[pattern-recognition#hough_transforms|[+] ]]
    * Standard Hough Transform
    * Randomized Hough Transform
    * Connective Randomized Hough Transform
    * Combinatorial Hough Transform
    * Adaptive Hough Transform
    * Probabilistic Hough Transform
    * Adaptive Probabilistic Hough Transform
    * Progressive Probabilistic Hough Transform
    * Hierarchical Hough Transform
    * Sampling Hough Transform
    * Generalized Hough Transform
  * UpWrite method
  * Curvogram
  * **Shape Descriptors**
    * Ankerst's Shape histograms
  * Chamfer matching
    * Contour Likelihood Measurement


=== Tracking ===

  * Kalman Filter
    * Generalized Kalman Filter
  * Correlation Tracking
    * ACA (Area Correlation Algorithm)
  * KLT Tracker (Kanade-Lucas-Tomasi)
  * IPAN Tracker
  * MeanShift
  * **Features detection**
    * Harris detector
    * Susan detector
    * Multiresolution Contrast detector

==== Sensors fusion ====

  * Kalman filter
  * Particles filter (bayesian network) - //filtrage particulaire//