SlideShare une entreprise Scribd logo
1  sur  25
Télécharger pour lire hors ligne
Foreground Detection via Robust Low Rank Matrix
       Factorization including Spatial Constraint with Iterative
                        Reweighted Regression

                             C. Guyon, T. Bouwmans and E. Zahzah
                                    charles.guyon@univ-lr.fr

       MIA Laboratory (Mathematics Images & Applications), University of La Rochelle, France
                  Presenter: Muriel Visani (L3i lab - University of la Rochelle)
                                                —
                                   ICPR2012, Tsukuba, Japan


                                            November 14, 2012



C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              1 Spatia
Summary



  1   Introduction and motivation on IRLS

  2   Temporal constraint with an adapted norm

  3   Diagram flow and spatial constraint

  4   Experimental Results

  5   Conclusion




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              2 Spatia
Introduction and motivation

   Purpose
          Foreground detection : Segmentation of moving objects in video sequence
          acquired by a fixed camera.
          Background modeling : Modelization of all that is not moving object.

   Involved applications
          Surveillance camera
          Motion capture

   On the importance
          Crucial Task : Often the first step of a full video surveillance system.

   Strategy used
          Eigenbackground decomposition.
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              3 Spatia
Eigenbackgrounds
          Find an « ideal » subspace of the video sequence, which describes the best
          as possible the (dynamic) background.




   Fig.1 The common process of background subtraction via PCA (Principal Component
   Analysis). At the final step, an adaptative threshold is used to get a binary image.

   Without a robust framework, the moving object may be absorbed in the model !
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              4 Spatia
Data Structure Transformation
   First, we consider a video sequence as a matrix A ∈ Rn×m




          n is the amount of pixels in a frame (∼ 106 )
          m is the number of frames considered (∼ 200)
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              5 Spatia
IRLS : Vector version (1)

   The usual IRLS (Iteratively Reweighted Least Squares) scheme for solve
   argmin ||Ax − b||α is given by :
        x

                                   D (i) = diag((ε + |b − Ax (i) |)α−2 )
                                                                                                            (1)
                                   x (i+1) = (At D (i) A)−1 At D (i) b

   This IRLS method is convergent for 1 ≤ α < 3. An more suitable
   formulation is :

                                      r (i) = b − Ax (i)
                                      D = diag((ε + |r (i) |)α−2 )
                                                                                                            (2)
                                      y (i) = (A DA)−1 A Dr (i)
                                      x (i+1) = x (i) + (1 + λopt )y (i)
   for λopt computed in a second inner loop. It is convergent for 1 ≤ α < +∞

C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              6 Spatia
IRLS : Vector version (2)


   For spatio/temporal RPCA, it needs to solve the following general problem :

                                argmin ||Ax − b||α + λ||Cx − d||β                                           (3)
                                     x

   By derivation, the associated IRLS scheme is,
                      r1 = b − Ax (i) , r2 = d − Cx (i) , e1 = ε + |r1 |, e2 = ε + |r2 |
                                    α 1                               β 1 −1         β−2
                      D1 = ( e1 ) α −1 diag(e1 ), D2 = λ( e2 ) β diag(e2 )
                                                  α−2
                        (i)                        −1
                                                                                                             (4)
                      y = (A D1 A + C D2 C ) (A D1 r1 + C D2 r2 )
                      x (i+1) = x (i) + (1 + λopt )y (i)



   Good news : Just few lines in Matlab !



C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              7 Spatia
IRLS : Matrix Version

   More generally, we consider the following matrix regression problem with
   two parameters norm (α, β) and a weighted matrix W ,
                                                                      n       m
                                                                                                 α   1
       min ||AX − B||α,β                with ||Mij ||α,β = (              (       Wij |Mij |β ) β ) α       (5)
         X                      W                          W
                                                                     i=1 j=1

  The problem is solved in the same manner on matrices with a reweighted
  regression strategy,
                      Until X   is stable, repeat on each k-column
                        R       ← B − AX
                         S      ← ε + |R|                                                                    (6)
                                                                 α −1
                                             β−2         β
                        Dk      ← diag(Sik ◦ ( j (Sij ◦ Wij )) β ◦ Wik )k
                        Xik     ← Xik +(1+Λ(max(α, β)))(At Dk A)−1 At Dk Rik




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              8 Spatia
Various RPCA formulation (only for α = 1)
          PCA with a fixed rank is :               min         ||S||F
                                                     L,S
                                                  s.t.        Rank(L) = k                                    (7)
                                                              A=L+S
          R(obust)PCA is (Non convex and NP-hard ) :
                                               min         ||σ(L)||0 + λ||S||0
                                               L,S                                                           (8)
                                               s.t.        A=L+S
          Convex relaxed problem of (8) is RPCA-PCP proposed by Candès et al. [1] :
                                               min         ||σ(L)||1 + λ||S||1
                                               L,S                                                           (9)
                                               s.t.        A=L+S
          where σ(L) means singular values of L.
          A mix is Stable PCP of Zhou et al. [2] (both entry-wise and sparse noise) :
                                              min          ||σ(L)||1 + λ||S||1
                                               L,S                                                          (10)
                                              s.t.         ||A − L − S||F < δ

          All of them could be solved by Augmented Lagrangian Multipliers (ALM).
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images2012
                                                                                   November 14, & Applications),
                                                                                                              9 Spatia
Video examples




   Some examples, temporal RPCA and ideal RPCA with ground truth fitting.




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          10 Spatia
Summary



  1   Introduction and motivation on IRLS

  2   Temporal constraint with an adapted norm

  3   Diagram flow and spatial constraint

  4   Experimental Results

  5   Conclusion




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          11 Spatia
Sparse solution
   In RPCA, residual error is sparse.
   Using the RPCA decomposition on a synthetic low-rank random matrix
   plus noise, the error looks like :




   Same principle with video. Sparse noise (or outliers) are the moving objects.
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          12 Spatia
Let’s play with norms




   Varying the α, β norm → Different kind of recovering pattern error.

C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          13 Spatia
Let’s play with norms...(2)




   Some issues
          What is the best specific norm for temporal constrain ?
          Initial assumption is ||.||2,1 . Confirmed experimentally ?




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          14 Spatia
Validation
   If ideal eigenbakgrounds are that, best norm should be ...




   Let us denote Lopt , the ideal low-rank subspace which outliers do not contribute to PCA
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          15 Spatia
Experimental validation

                                                            Let us denote Lα,β , the low-rank
                                                            recovered matrix with a ||.||α,β -PCA.
                                                            The plot shows the error between
                                                            ||Lopt − Lα,β ||F for parameters α
                                                            and β chosen freely. The darkest
                                                            value means that the error is the
                                                            smallest here.


          ||S||2,1 is not optimal, but for convenience we use it.
          The benefit of the ad hoc block-sparse hypothesis is confirmed by
          testing its efficiency directly on video dataset.

         Experimentation done on dynamic category of dataset change detection
                workshop 2012 : http://www.changedetection.net/

C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          16 Spatia
Summary



  1   Introduction and motivation on IRLS

  2   Temporal constraint with an adapted norm

  3   Diagram flow and spatial constraint

  4   Experimental Results

  5   Conclusion




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          17 Spatia
Overview & addition of a spatial constraint via TV




   Figure: Overview of the learning and evaluation process. Learning process needs
   GT (Ground Truth) for better fits the eigenbackground components.


   Spatial Constraint via TV
          Suppose A = L + S where L and S are computed via some kind of
          RPCA techniques with the addition of Total Variation penalty on S.
          This penalty increases connected (or connexe) shapes.

C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          18 Spatia
Exemple with a synthetic 1-D signal




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          19 Spatia
Summary



  1   Introduction and motivation on IRLS

  2   Temporal constraint with an adapted norm

  3   Diagram flow and spatial constraint

  4   Experimental Results

  5   Conclusion




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          20 Spatia
Experimental Protocol
   RPCA-IRLS is compared for the following four recent robust methods :
          Low-Rank Block sparse Decomposition (LBD, 2011) [3]
          Low-Rank Representation (LRR, 2011) [4]
          Symmetric Alternating Direction Augmented Lagrangian (SADAL, 2011) [5]
          Grassmannian Robust Adaptive Subspace Tracking Algorithm (GRASTA, 2012) [6]

   References
      [1] E. Candes, X. Li, Y. Ma, and J. Wright, Robust principal component analysis, International Journal of ACM,
          vol. 58, no. 3, May 2011.
      [2] Z. Zhou, X. Li, J. Wright, E. Candes, and Y. Ma, Stable principal component pursuit,IEEE ISIT
          Proceedings, pp. 1518-1522, Jun. 2010.
      [3] G. Tang and A. Nehorai, Robust principal component analysis based on low-rank and block-sparse matrix
          decomposition, CISS 2011, 2011.
      [4] Z. Lin, R. Liu, and Z. Su. Linearized alternating direction method with adaptive penalty for low-rank
          representation. NIPS 2011, Dec. 2011.
      [5] S. Ma. Algorithms for sparse and low-rank optimization : Convergence, complexity and applications. Thesis,
          2011.
      [6] J. He, L. Balzano, and A. Szlam. Incremental gradient on the grassmannian for online foreground and
          background separation in subsampled video. Conference on Computer Vision and Pattern Recognition
          (CVPR), June 2012.



C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          21 Spatia
Experimental Protocol & Quantitative Results

          Optimal threshold is chosen for maximizing F-measure criterion
          which is based 2 × 2 histogram of True/false/positive/negative :

                                TP                             TP                     2 DR Prec
                    DR =              ,          Prec =              ,         F =
                              TP + FN                        TP + FP                  DR + Prec


          Good performance is then obtained when the F-measure is closed to 1
          Time consumption is not take into account in the evaluation process.




                      Figure: F-Measure on the Wallflower and I2R dataset.


C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          22 Spatia
Quantitative Results

   Here, we show other experimental results on the real dataset of BMC 2012,


      Video       Recall        Precision         F-measure          PSNR               Visual Results
        1         0.9139         0.7170             0.8036           38.2425
        2         0.8785         0.8656             0.8720           26.7721
        3         0.9658         0.8120             0.8822           37.7053
        4         0.9550         0.7187             0.8202           39.3699
        5         0.9102         0.5589             0.6925           30.5876
        6         0.9002         0.7727             0.8316           29.9994
        7         0.9116         0.8401             0.8744           26.8350
        8         0.8651         0.6710             0.7558           30.5040
        9         0.9309         0.8239             0.8741           55.1163

   Table: Quantitative results with common criterions. Last column : sample of the
   original video, GT and our results of the first four real video sequences.


C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          23 Spatia
Summary



  1   Introduction and motivation on IRLS

  2   Temporal constraint with an adapted norm

  3   Diagram flow and spatial constraint

  4   Experimental Results

  5   Conclusion




C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          24 Spatia
Conclusion

   Advantages
          Experiments on video surveillance datasets show that this approach is more
          robust than other recent RPCA formulation in presence of dynamic
          backgrounds (DC) and illumination changes (IC).
          Well suited for video with spatially spread and temporarily sparse outliers.

   Disadvantages
          Small local motions, like « waving trees » are not (yet) well modelized by
          this kind of global PCA. For example, IC needs few eigenBackground and
          DC needs more with the risk to integrate moving objects into the model.

   Future Works
          Lack in computation time : Further research consists in developping an
          incremental version to update the model at every frame and to achieve the
          real-time requirements.
C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U
                                     Foreground Detection (MIA Laboratory (Mathematics Images & Applications),
                                                                                  November 14, 2012          25 Spatia

Contenu connexe

Tendances

Robust Image Denoising in RKHS via Orthogonal Matching Pursuit
Robust Image Denoising in RKHS via Orthogonal Matching PursuitRobust Image Denoising in RKHS via Orthogonal Matching Pursuit
Robust Image Denoising in RKHS via Orthogonal Matching PursuitPantelis Bouboulis
 
Continuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDContinuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDValentin De Bortoli
 
From RNN to neural networks for cyclic undirected graphs
From RNN to neural networks for cyclic undirected graphsFrom RNN to neural networks for cyclic undirected graphs
From RNN to neural networks for cyclic undirected graphstuxette
 
Dictionary Learning for Massive Matrix Factorization
Dictionary Learning for Massive Matrix FactorizationDictionary Learning for Massive Matrix Factorization
Dictionary Learning for Massive Matrix Factorizationrecsysfr
 
Data-Driven Recommender Systems
Data-Driven Recommender SystemsData-Driven Recommender Systems
Data-Driven Recommender Systemsrecsysfr
 
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...Dongmin Choi
 
MLIP - Chapter 2 - Preliminaries to deep learning
MLIP - Chapter 2 - Preliminaries to deep learningMLIP - Chapter 2 - Preliminaries to deep learning
MLIP - Chapter 2 - Preliminaries to deep learningCharles Deledalle
 
Particle Filters and Applications in Computer Vision
Particle Filters and Applications in Computer VisionParticle Filters and Applications in Computer Vision
Particle Filters and Applications in Computer Visionzukun
 
A short and naive introduction to using network in prediction models
A short and naive introduction to using network in prediction modelsA short and naive introduction to using network in prediction models
A short and naive introduction to using network in prediction modelstuxette
 
Csss2010 20100803-kanevski-lecture2
Csss2010 20100803-kanevski-lecture2Csss2010 20100803-kanevski-lecture2
Csss2010 20100803-kanevski-lecture2hasan_elektro
 
Habilitation à diriger des recherches
Habilitation à diriger des recherchesHabilitation à diriger des recherches
Habilitation à diriger des recherchesPierre Pudlo
 
Wireless Localization: Positioning
Wireless Localization: PositioningWireless Localization: Positioning
Wireless Localization: PositioningStefano Severi
 
Expert Lecture on GPS at UIET, CSJM, Kanpur
Expert Lecture on GPS at UIET, CSJM, KanpurExpert Lecture on GPS at UIET, CSJM, Kanpur
Expert Lecture on GPS at UIET, CSJM, KanpurSuddhasheel GHOSH, PhD
 
Global illumination techniques for the computation of hight quality images in...
Global illumination techniques for the computation of hight quality images in...Global illumination techniques for the computation of hight quality images in...
Global illumination techniques for the computation of hight quality images in...Frederic Perez
 
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...Yusuf Bhujwalla
 
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Dongmin Choi
 

Tendances (20)

Bayesian Core: Chapter 8
Bayesian Core: Chapter 8Bayesian Core: Chapter 8
Bayesian Core: Chapter 8
 
Robust Image Denoising in RKHS via Orthogonal Matching Pursuit
Robust Image Denoising in RKHS via Orthogonal Matching PursuitRobust Image Denoising in RKHS via Orthogonal Matching Pursuit
Robust Image Denoising in RKHS via Orthogonal Matching Pursuit
 
Continuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDContinuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGD
 
From RNN to neural networks for cyclic undirected graphs
From RNN to neural networks for cyclic undirected graphsFrom RNN to neural networks for cyclic undirected graphs
From RNN to neural networks for cyclic undirected graphs
 
Dictionary Learning for Massive Matrix Factorization
Dictionary Learning for Massive Matrix FactorizationDictionary Learning for Massive Matrix Factorization
Dictionary Learning for Massive Matrix Factorization
 
Data-Driven Recommender Systems
Data-Driven Recommender SystemsData-Driven Recommender Systems
Data-Driven Recommender Systems
 
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
 
Pres metabief2020jmm
Pres metabief2020jmmPres metabief2020jmm
Pres metabief2020jmm
 
MLIP - Chapter 2 - Preliminaries to deep learning
MLIP - Chapter 2 - Preliminaries to deep learningMLIP - Chapter 2 - Preliminaries to deep learning
MLIP - Chapter 2 - Preliminaries to deep learning
 
Presentation v3.2
Presentation v3.2Presentation v3.2
Presentation v3.2
 
Particle Filters and Applications in Computer Vision
Particle Filters and Applications in Computer VisionParticle Filters and Applications in Computer Vision
Particle Filters and Applications in Computer Vision
 
A short and naive introduction to using network in prediction models
A short and naive introduction to using network in prediction modelsA short and naive introduction to using network in prediction models
A short and naive introduction to using network in prediction models
 
Csss2010 20100803-kanevski-lecture2
Csss2010 20100803-kanevski-lecture2Csss2010 20100803-kanevski-lecture2
Csss2010 20100803-kanevski-lecture2
 
Habilitation à diriger des recherches
Habilitation à diriger des recherchesHabilitation à diriger des recherches
Habilitation à diriger des recherches
 
Wireless Localization: Positioning
Wireless Localization: PositioningWireless Localization: Positioning
Wireless Localization: Positioning
 
CSMR11b.ppt
CSMR11b.pptCSMR11b.ppt
CSMR11b.ppt
 
Expert Lecture on GPS at UIET, CSJM, Kanpur
Expert Lecture on GPS at UIET, CSJM, KanpurExpert Lecture on GPS at UIET, CSJM, Kanpur
Expert Lecture on GPS at UIET, CSJM, Kanpur
 
Global illumination techniques for the computation of hight quality images in...
Global illumination techniques for the computation of hight quality images in...Global illumination techniques for the computation of hight quality images in...
Global illumination techniques for the computation of hight quality images in...
 
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...
The Impact of Smoothness on Model Class Selection in Nonlinear System Identif...
 
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
 

En vedette

Yannick Chevalier - Habilitation (final)
Yannick Chevalier - Habilitation (final)Yannick Chevalier - Habilitation (final)
Yannick Chevalier - Habilitation (final)Yannick Chevalier
 
Habilitation presentation 2011
Habilitation presentation 2011Habilitation presentation 2011
Habilitation presentation 2011Veronica Bermudez
 
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)Vincent Fetzer
 
Presentation Habilitation à Diriger des Recherches
Presentation Habilitation à Diriger des RecherchesPresentation Habilitation à Diriger des Recherches
Presentation Habilitation à Diriger des RecherchesBOUWMANS Thierry
 
Non-negative factorization methods for extracting semantically relevant featu...
Non-negative factorization methods for extracting semantically relevant featu...Non-negative factorization methods for extracting semantically relevant featu...
Non-negative factorization methods for extracting semantically relevant featu...Gabriella Casalino
 
Financial methods in online advertising
 Financial methods in online advertising Financial methods in online advertising
Financial methods in online advertisingJun Wang
 
Incremental and Multi-feature Tensor Subspace Learning applied for Background...
Incremental and Multi-feature Tensor Subspace Learning applied for Background...Incremental and Multi-feature Tensor Subspace Learning applied for Background...
Incremental and Multi-feature Tensor Subspace Learning applied for Background...ActiveEon
 
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...Vasily Leksin
 
Warsaw Data Science - Factorization Machines Introduction
Warsaw Data Science -  Factorization Machines IntroductionWarsaw Data Science -  Factorization Machines Introduction
Warsaw Data Science - Factorization Machines IntroductionBartlomiej Twardowski
 
(2016 07-19) providing click predictions in real-time at scale
(2016 07-19) providing click predictions in real-time at scale(2016 07-19) providing click predictions in real-time at scale
(2016 07-19) providing click predictions in real-time at scaleLawrence Evans
 
Intro to Factorization Machines
Intro to Factorization MachinesIntro to Factorization Machines
Intro to Factorization MachinesPavel Kalaidin
 
Exploiting Ranking Factorization Machines for Microblog Retrieval
Exploiting Ranking Factorization Machines for Microblog RetrievalExploiting Ranking Factorization Machines for Microblog Retrieval
Exploiting Ranking Factorization Machines for Microblog RetrievalRunwei Qiang
 
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...Shuai Yuan
 
Factorization Machines with libFM
Factorization Machines with libFMFactorization Machines with libFM
Factorization Machines with libFMLiangjie Hong
 
CORE: Context-Aware Open Relation Extraction with Factorization Machines
CORE: Context-Aware Open Relation Extraction with Factorization MachinesCORE: Context-Aware Open Relation Extraction with Factorization Machines
CORE: Context-Aware Open Relation Extraction with Factorization MachinesFabio Petroni, PhD
 
Steffen Rendle, Research Scientist, Google at MLconf SF
Steffen Rendle, Research Scientist, Google at MLconf SFSteffen Rendle, Research Scientist, Google at MLconf SF
Steffen Rendle, Research Scientist, Google at MLconf SFMLconf
 
ESSIR 2013 Recommender Systems tutorial
ESSIR 2013 Recommender Systems tutorial ESSIR 2013 Recommender Systems tutorial
ESSIR 2013 Recommender Systems tutorial Alexandros Karatzoglou
 
RTBMA ECIR 2016 tutorial
RTBMA ECIR 2016 tutorialRTBMA ECIR 2016 tutorial
RTBMA ECIR 2016 tutorialShuai Yuan
 

En vedette (20)

Yannick Chevalier - Habilitation (final)
Yannick Chevalier - Habilitation (final)Yannick Chevalier - Habilitation (final)
Yannick Chevalier - Habilitation (final)
 
Habilitation presentation 2011
Habilitation presentation 2011Habilitation presentation 2011
Habilitation presentation 2011
 
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)
Habilitation électrique Vincent Fetzer (BT : BC, BR, B2V essais)
 
Presentation Habilitation à Diriger des Recherches
Presentation Habilitation à Diriger des RecherchesPresentation Habilitation à Diriger des Recherches
Presentation Habilitation à Diriger des Recherches
 
Non-negative factorization methods for extracting semantically relevant featu...
Non-negative factorization methods for extracting semantically relevant featu...Non-negative factorization methods for extracting semantically relevant featu...
Non-negative factorization methods for extracting semantically relevant featu...
 
My PhD trajectory
My PhD trajectoryMy PhD trajectory
My PhD trajectory
 
Financial methods in online advertising
 Financial methods in online advertising Financial methods in online advertising
Financial methods in online advertising
 
Thesis_Nazarova_Final(1)
Thesis_Nazarova_Final(1)Thesis_Nazarova_Final(1)
Thesis_Nazarova_Final(1)
 
Incremental and Multi-feature Tensor Subspace Learning applied for Background...
Incremental and Multi-feature Tensor Subspace Learning applied for Background...Incremental and Multi-feature Tensor Subspace Learning applied for Background...
Incremental and Multi-feature Tensor Subspace Learning applied for Background...
 
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...
Avito recsys-challenge-2016RecSys Challenge 2016: Job Recommendation Based on...
 
Warsaw Data Science - Factorization Machines Introduction
Warsaw Data Science -  Factorization Machines IntroductionWarsaw Data Science -  Factorization Machines Introduction
Warsaw Data Science - Factorization Machines Introduction
 
(2016 07-19) providing click predictions in real-time at scale
(2016 07-19) providing click predictions in real-time at scale(2016 07-19) providing click predictions in real-time at scale
(2016 07-19) providing click predictions in real-time at scale
 
Intro to Factorization Machines
Intro to Factorization MachinesIntro to Factorization Machines
Intro to Factorization Machines
 
Exploiting Ranking Factorization Machines for Microblog Retrieval
Exploiting Ranking Factorization Machines for Microblog RetrievalExploiting Ranking Factorization Machines for Microblog Retrieval
Exploiting Ranking Factorization Machines for Microblog Retrieval
 
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...
CIKM 2013 Tutorial: Real-time Bidding: A New Frontier of Computational Advert...
 
Factorization Machines with libFM
Factorization Machines with libFMFactorization Machines with libFM
Factorization Machines with libFM
 
CORE: Context-Aware Open Relation Extraction with Factorization Machines
CORE: Context-Aware Open Relation Extraction with Factorization MachinesCORE: Context-Aware Open Relation Extraction with Factorization Machines
CORE: Context-Aware Open Relation Extraction with Factorization Machines
 
Steffen Rendle, Research Scientist, Google at MLconf SF
Steffen Rendle, Research Scientist, Google at MLconf SFSteffen Rendle, Research Scientist, Google at MLconf SF
Steffen Rendle, Research Scientist, Google at MLconf SF
 
ESSIR 2013 Recommender Systems tutorial
ESSIR 2013 Recommender Systems tutorial ESSIR 2013 Recommender Systems tutorial
ESSIR 2013 Recommender Systems tutorial
 
RTBMA ECIR 2016 tutorial
RTBMA ECIR 2016 tutorialRTBMA ECIR 2016 tutorial
RTBMA ECIR 2016 tutorial
 

Similaire à ICPR 2012

Digital image processing short quesstion answers
Digital image processing short quesstion answersDigital image processing short quesstion answers
Digital image processing short quesstion answersAteeq Zada
 
Projection methods for stochastic structural dynamics
Projection methods for stochastic structural dynamicsProjection methods for stochastic structural dynamics
Projection methods for stochastic structural dynamicsUniversity of Glasgow
 
Fast Object Recognition from 3D Depth Data with Extreme Learning Machine
Fast Object Recognition from 3D Depth Data with Extreme Learning MachineFast Object Recognition from 3D Depth Data with Extreme Learning Machine
Fast Object Recognition from 3D Depth Data with Extreme Learning MachineSoma Boubou
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysisbutest
 
Large scale landuse classification of satellite imagery
Large scale landuse classification of satellite imageryLarge scale landuse classification of satellite imagery
Large scale landuse classification of satellite imagerySuneel Marthi
 
Landuse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningLanduse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningDataWorks Summit
 
Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Valentin De Bortoli
 
Linear Smoothing, Median, and Sharpening Filter
Linear Smoothing, Median, and Sharpening FilterLinear Smoothing, Median, and Sharpening Filter
Linear Smoothing, Median, and Sharpening FilterVARUN KUMAR
 
Iaetsd vlsi implementation of gabor filter based image edge detection
Iaetsd vlsi implementation of gabor filter based image edge detectionIaetsd vlsi implementation of gabor filter based image edge detection
Iaetsd vlsi implementation of gabor filter based image edge detectionIaetsd Iaetsd
 
Sequential Extraction of Local ICA Structures
Sequential Extraction of Local ICA StructuresSequential Extraction of Local ICA Structures
Sequential Extraction of Local ICA Structurestopujahin
 
机器学习Adaboost
机器学习Adaboost机器学习Adaboost
机器学习AdaboostShocky1
 

Similaire à ICPR 2012 (20)

Foreground Detection : Combining Background Subspace Learning with Object Smo...
Foreground Detection : Combining Background Subspace Learning with Object Smo...Foreground Detection : Combining Background Subspace Learning with Object Smo...
Foreground Detection : Combining Background Subspace Learning with Object Smo...
 
Digital image processing short quesstion answers
Digital image processing short quesstion answersDigital image processing short quesstion answers
Digital image processing short quesstion answers
 
Projection methods for stochastic structural dynamics
Projection methods for stochastic structural dynamicsProjection methods for stochastic structural dynamics
Projection methods for stochastic structural dynamics
 
Fast Object Recognition from 3D Depth Data with Extreme Learning Machine
Fast Object Recognition from 3D Depth Data with Extreme Learning MachineFast Object Recognition from 3D Depth Data with Extreme Learning Machine
Fast Object Recognition from 3D Depth Data with Extreme Learning Machine
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Machine Learning and Statistical Analysis
Machine Learning and Statistical AnalysisMachine Learning and Statistical Analysis
Machine Learning and Statistical Analysis
 
Large scale landuse classification of satellite imagery
Large scale landuse classification of satellite imageryLarge scale landuse classification of satellite imagery
Large scale landuse classification of satellite imagery
 
Landuse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningLanduse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep Learning
 
ABC workshop: 17w5025
ABC workshop: 17w5025ABC workshop: 17w5025
ABC workshop: 17w5025
 
Lec-3 DIP.pptx
Lec-3 DIP.pptxLec-3 DIP.pptx
Lec-3 DIP.pptx
 
Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...
 
Linear Smoothing, Median, and Sharpening Filter
Linear Smoothing, Median, and Sharpening FilterLinear Smoothing, Median, and Sharpening Filter
Linear Smoothing, Median, and Sharpening Filter
 
Iaetsd vlsi implementation of gabor filter based image edge detection
Iaetsd vlsi implementation of gabor filter based image edge detectionIaetsd vlsi implementation of gabor filter based image edge detection
Iaetsd vlsi implementation of gabor filter based image edge detection
 
Sequential Extraction of Local ICA Structures
Sequential Extraction of Local ICA StructuresSequential Extraction of Local ICA Structures
Sequential Extraction of Local ICA Structures
 
机器学习Adaboost
机器学习Adaboost机器学习Adaboost
机器学习Adaboost
 

Dernier

How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyes
How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyesHow to Effectively Monitor SD-WAN and SASE Environments with ThousandEyes
How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyesThousandEyes
 
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...Jeffrey Haguewood
 
QCon London: Mastering long-running processes in modern architectures
QCon London: Mastering long-running processes in modern architecturesQCon London: Mastering long-running processes in modern architectures
QCon London: Mastering long-running processes in modern architecturesBernd Ruecker
 
WomenInAutomation2024: AI and Automation for eveyone
WomenInAutomation2024: AI and Automation for eveyoneWomenInAutomation2024: AI and Automation for eveyone
WomenInAutomation2024: AI and Automation for eveyoneUiPathCommunity
 
A Glance At The Java Performance Toolbox
A Glance At The Java Performance ToolboxA Glance At The Java Performance Toolbox
A Glance At The Java Performance ToolboxAna-Maria Mihalceanu
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPathCommunity
 
A Framework for Development in the AI Age
A Framework for Development in the AI AgeA Framework for Development in the AI Age
A Framework for Development in the AI AgeCprime
 
Potential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsPotential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsRavi Sanghani
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesKari Kakkonen
 
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24Mark Goldstein
 
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...Nikki Chapple
 
Glenn Lazarus- Why Your Observability Strategy Needs Security Observability
Glenn Lazarus- Why Your Observability Strategy Needs Security ObservabilityGlenn Lazarus- Why Your Observability Strategy Needs Security Observability
Glenn Lazarus- Why Your Observability Strategy Needs Security Observabilityitnewsafrica
 
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality AssuranceInflectra
 
Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Farhan Tariq
 
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...panagenda
 
Manual 508 Accessibility Compliance Audit
Manual 508 Accessibility Compliance AuditManual 508 Accessibility Compliance Audit
Manual 508 Accessibility Compliance AuditSkynet Technologies
 
All These Sophisticated Attacks, Can We Really Detect Them - PDF
All These Sophisticated Attacks, Can We Really Detect Them - PDFAll These Sophisticated Attacks, Can We Really Detect Them - PDF
All These Sophisticated Attacks, Can We Really Detect Them - PDFMichael Gough
 
Assure Ecommerce and Retail Operations Uptime with ThousandEyes
Assure Ecommerce and Retail Operations Uptime with ThousandEyesAssure Ecommerce and Retail Operations Uptime with ThousandEyes
Assure Ecommerce and Retail Operations Uptime with ThousandEyesThousandEyes
 
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...JET Technology Labs White Paper for Virtualized Security and Encryption Techn...
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...amber724300
 
Connecting the Dots for Information Discovery.pdf
Connecting the Dots for Information Discovery.pdfConnecting the Dots for Information Discovery.pdf
Connecting the Dots for Information Discovery.pdfNeo4j
 

Dernier (20)

How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyes
How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyesHow to Effectively Monitor SD-WAN and SASE Environments with ThousandEyes
How to Effectively Monitor SD-WAN and SASE Environments with ThousandEyes
 
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...
Email Marketing Automation for Bonterra Impact Management (fka Social Solutio...
 
QCon London: Mastering long-running processes in modern architectures
QCon London: Mastering long-running processes in modern architecturesQCon London: Mastering long-running processes in modern architectures
QCon London: Mastering long-running processes in modern architectures
 
WomenInAutomation2024: AI and Automation for eveyone
WomenInAutomation2024: AI and Automation for eveyoneWomenInAutomation2024: AI and Automation for eveyone
WomenInAutomation2024: AI and Automation for eveyone
 
A Glance At The Java Performance Toolbox
A Glance At The Java Performance ToolboxA Glance At The Java Performance Toolbox
A Glance At The Java Performance Toolbox
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to Hero
 
A Framework for Development in the AI Age
A Framework for Development in the AI AgeA Framework for Development in the AI Age
A Framework for Development in the AI Age
 
Potential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsPotential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and Insights
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examples
 
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24
Arizona Broadband Policy Past, Present, and Future Presentation 3/25/24
 
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...
Microsoft 365 Copilot: How to boost your productivity with AI – Part one: Ado...
 
Glenn Lazarus- Why Your Observability Strategy Needs Security Observability
Glenn Lazarus- Why Your Observability Strategy Needs Security ObservabilityGlenn Lazarus- Why Your Observability Strategy Needs Security Observability
Glenn Lazarus- Why Your Observability Strategy Needs Security Observability
 
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
 
Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...
 
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
 
Manual 508 Accessibility Compliance Audit
Manual 508 Accessibility Compliance AuditManual 508 Accessibility Compliance Audit
Manual 508 Accessibility Compliance Audit
 
All These Sophisticated Attacks, Can We Really Detect Them - PDF
All These Sophisticated Attacks, Can We Really Detect Them - PDFAll These Sophisticated Attacks, Can We Really Detect Them - PDF
All These Sophisticated Attacks, Can We Really Detect Them - PDF
 
Assure Ecommerce and Retail Operations Uptime with ThousandEyes
Assure Ecommerce and Retail Operations Uptime with ThousandEyesAssure Ecommerce and Retail Operations Uptime with ThousandEyes
Assure Ecommerce and Retail Operations Uptime with ThousandEyes
 
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...JET Technology Labs White Paper for Virtualized Security and Encryption Techn...
JET Technology Labs White Paper for Virtualized Security and Encryption Techn...
 
Connecting the Dots for Information Discovery.pdf
Connecting the Dots for Information Discovery.pdfConnecting the Dots for Information Discovery.pdf
Connecting the Dots for Information Discovery.pdf
 

ICPR 2012

  • 1. Foreground Detection via Robust Low Rank Matrix Factorization including Spatial Constraint with Iterative Reweighted Regression C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr MIA Laboratory (Mathematics Images & Applications), University of La Rochelle, France Presenter: Muriel Visani (L3i lab - University of la Rochelle) — ICPR2012, Tsukuba, Japan November 14, 2012 C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 1 Spatia
  • 2. Summary 1 Introduction and motivation on IRLS 2 Temporal constraint with an adapted norm 3 Diagram flow and spatial constraint 4 Experimental Results 5 Conclusion C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 2 Spatia
  • 3. Introduction and motivation Purpose Foreground detection : Segmentation of moving objects in video sequence acquired by a fixed camera. Background modeling : Modelization of all that is not moving object. Involved applications Surveillance camera Motion capture On the importance Crucial Task : Often the first step of a full video surveillance system. Strategy used Eigenbackground decomposition. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 3 Spatia
  • 4. Eigenbackgrounds Find an « ideal » subspace of the video sequence, which describes the best as possible the (dynamic) background. Fig.1 The common process of background subtraction via PCA (Principal Component Analysis). At the final step, an adaptative threshold is used to get a binary image. Without a robust framework, the moving object may be absorbed in the model ! C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 4 Spatia
  • 5. Data Structure Transformation First, we consider a video sequence as a matrix A ∈ Rn×m n is the amount of pixels in a frame (∼ 106 ) m is the number of frames considered (∼ 200) C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 5 Spatia
  • 6. IRLS : Vector version (1) The usual IRLS (Iteratively Reweighted Least Squares) scheme for solve argmin ||Ax − b||α is given by : x D (i) = diag((ε + |b − Ax (i) |)α−2 ) (1) x (i+1) = (At D (i) A)−1 At D (i) b This IRLS method is convergent for 1 ≤ α < 3. An more suitable formulation is : r (i) = b − Ax (i) D = diag((ε + |r (i) |)α−2 ) (2) y (i) = (A DA)−1 A Dr (i) x (i+1) = x (i) + (1 + λopt )y (i) for λopt computed in a second inner loop. It is convergent for 1 ≤ α < +∞ C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 6 Spatia
  • 7. IRLS : Vector version (2) For spatio/temporal RPCA, it needs to solve the following general problem : argmin ||Ax − b||α + λ||Cx − d||β (3) x By derivation, the associated IRLS scheme is, r1 = b − Ax (i) , r2 = d − Cx (i) , e1 = ε + |r1 |, e2 = ε + |r2 | α 1 β 1 −1 β−2 D1 = ( e1 ) α −1 diag(e1 ), D2 = λ( e2 ) β diag(e2 ) α−2 (i) −1 (4) y = (A D1 A + C D2 C ) (A D1 r1 + C D2 r2 ) x (i+1) = x (i) + (1 + λopt )y (i) Good news : Just few lines in Matlab ! C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 7 Spatia
  • 8. IRLS : Matrix Version More generally, we consider the following matrix regression problem with two parameters norm (α, β) and a weighted matrix W , n m α 1 min ||AX − B||α,β with ||Mij ||α,β = ( ( Wij |Mij |β ) β ) α (5) X W W i=1 j=1 The problem is solved in the same manner on matrices with a reweighted regression strategy, Until X is stable, repeat on each k-column R ← B − AX S ← ε + |R| (6) α −1 β−2 β Dk ← diag(Sik ◦ ( j (Sij ◦ Wij )) β ◦ Wik )k Xik ← Xik +(1+Λ(max(α, β)))(At Dk A)−1 At Dk Rik C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 8 Spatia
  • 9. Various RPCA formulation (only for α = 1) PCA with a fixed rank is : min ||S||F L,S s.t. Rank(L) = k (7) A=L+S R(obust)PCA is (Non convex and NP-hard ) : min ||σ(L)||0 + λ||S||0 L,S (8) s.t. A=L+S Convex relaxed problem of (8) is RPCA-PCP proposed by Candès et al. [1] : min ||σ(L)||1 + λ||S||1 L,S (9) s.t. A=L+S where σ(L) means singular values of L. A mix is Stable PCP of Zhou et al. [2] (both entry-wise and sparse noise) : min ||σ(L)||1 + λ||S||1 L,S (10) s.t. ||A − L − S||F < δ All of them could be solved by Augmented Lagrangian Multipliers (ALM). C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images2012 November 14, & Applications), 9 Spatia
  • 10. Video examples Some examples, temporal RPCA and ideal RPCA with ground truth fitting. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 10 Spatia
  • 11. Summary 1 Introduction and motivation on IRLS 2 Temporal constraint with an adapted norm 3 Diagram flow and spatial constraint 4 Experimental Results 5 Conclusion C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 11 Spatia
  • 12. Sparse solution In RPCA, residual error is sparse. Using the RPCA decomposition on a synthetic low-rank random matrix plus noise, the error looks like : Same principle with video. Sparse noise (or outliers) are the moving objects. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 12 Spatia
  • 13. Let’s play with norms Varying the α, β norm → Different kind of recovering pattern error. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 13 Spatia
  • 14. Let’s play with norms...(2) Some issues What is the best specific norm for temporal constrain ? Initial assumption is ||.||2,1 . Confirmed experimentally ? C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 14 Spatia
  • 15. Validation If ideal eigenbakgrounds are that, best norm should be ... Let us denote Lopt , the ideal low-rank subspace which outliers do not contribute to PCA C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 15 Spatia
  • 16. Experimental validation Let us denote Lα,β , the low-rank recovered matrix with a ||.||α,β -PCA. The plot shows the error between ||Lopt − Lα,β ||F for parameters α and β chosen freely. The darkest value means that the error is the smallest here. ||S||2,1 is not optimal, but for convenience we use it. The benefit of the ad hoc block-sparse hypothesis is confirmed by testing its efficiency directly on video dataset. Experimentation done on dynamic category of dataset change detection workshop 2012 : http://www.changedetection.net/ C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 16 Spatia
  • 17. Summary 1 Introduction and motivation on IRLS 2 Temporal constraint with an adapted norm 3 Diagram flow and spatial constraint 4 Experimental Results 5 Conclusion C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 17 Spatia
  • 18. Overview & addition of a spatial constraint via TV Figure: Overview of the learning and evaluation process. Learning process needs GT (Ground Truth) for better fits the eigenbackground components. Spatial Constraint via TV Suppose A = L + S where L and S are computed via some kind of RPCA techniques with the addition of Total Variation penalty on S. This penalty increases connected (or connexe) shapes. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 18 Spatia
  • 19. Exemple with a synthetic 1-D signal C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 19 Spatia
  • 20. Summary 1 Introduction and motivation on IRLS 2 Temporal constraint with an adapted norm 3 Diagram flow and spatial constraint 4 Experimental Results 5 Conclusion C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 20 Spatia
  • 21. Experimental Protocol RPCA-IRLS is compared for the following four recent robust methods : Low-Rank Block sparse Decomposition (LBD, 2011) [3] Low-Rank Representation (LRR, 2011) [4] Symmetric Alternating Direction Augmented Lagrangian (SADAL, 2011) [5] Grassmannian Robust Adaptive Subspace Tracking Algorithm (GRASTA, 2012) [6] References [1] E. Candes, X. Li, Y. Ma, and J. Wright, Robust principal component analysis, International Journal of ACM, vol. 58, no. 3, May 2011. [2] Z. Zhou, X. Li, J. Wright, E. Candes, and Y. Ma, Stable principal component pursuit,IEEE ISIT Proceedings, pp. 1518-1522, Jun. 2010. [3] G. Tang and A. Nehorai, Robust principal component analysis based on low-rank and block-sparse matrix decomposition, CISS 2011, 2011. [4] Z. Lin, R. Liu, and Z. Su. Linearized alternating direction method with adaptive penalty for low-rank representation. NIPS 2011, Dec. 2011. [5] S. Ma. Algorithms for sparse and low-rank optimization : Convergence, complexity and applications. Thesis, 2011. [6] J. He, L. Balzano, and A. Szlam. Incremental gradient on the grassmannian for online foreground and background separation in subsampled video. Conference on Computer Vision and Pattern Recognition (CVPR), June 2012. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 21 Spatia
  • 22. Experimental Protocol & Quantitative Results Optimal threshold is chosen for maximizing F-measure criterion which is based 2 × 2 histogram of True/false/positive/negative : TP TP 2 DR Prec DR = , Prec = , F = TP + FN TP + FP DR + Prec Good performance is then obtained when the F-measure is closed to 1 Time consumption is not take into account in the evaluation process. Figure: F-Measure on the Wallflower and I2R dataset. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 22 Spatia
  • 23. Quantitative Results Here, we show other experimental results on the real dataset of BMC 2012, Video Recall Precision F-measure PSNR Visual Results 1 0.9139 0.7170 0.8036 38.2425 2 0.8785 0.8656 0.8720 26.7721 3 0.9658 0.8120 0.8822 37.7053 4 0.9550 0.7187 0.8202 39.3699 5 0.9102 0.5589 0.6925 30.5876 6 0.9002 0.7727 0.8316 29.9994 7 0.9116 0.8401 0.8744 26.8350 8 0.8651 0.6710 0.7558 30.5040 9 0.9309 0.8239 0.8741 55.1163 Table: Quantitative results with common criterions. Last column : sample of the original video, GT and our results of the first four real video sequences. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 23 Spatia
  • 24. Summary 1 Introduction and motivation on IRLS 2 Temporal constraint with an adapted norm 3 Diagram flow and spatial constraint 4 Experimental Results 5 Conclusion C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 24 Spatia
  • 25. Conclusion Advantages Experiments on video surveillance datasets show that this approach is more robust than other recent RPCA formulation in presence of dynamic backgrounds (DC) and illumination changes (IC). Well suited for video with spatially spread and temporarily sparse outliers. Disadvantages Small local motions, like « waving trees » are not (yet) well modelized by this kind of global PCA. For example, IC needs few eigenBackground and DC needs more with the risk to integrate moving objects into the model. Future Works Lack in computation time : Further research consists in developping an incremental version to update the model at every frame and to achieve the real-time requirements. C. Guyon, T. Bouwmans and E. Zahzah charles.guyon@univ-lr.fr via Robust Low Rank Matrix Factorization including / 25 U Foreground Detection (MIA Laboratory (Mathematics Images & Applications), November 14, 2012 25 Spatia