Digit Recognition using SVMS

Slides:



Advertisements
Similar presentations
Max-Margin Additive Classifiers for Detection
Advertisements

Classification using intersection kernel SVMs is efficient
Jan-Michael Frahm, Enrique Dunn Spring 2013
Histograms of Oriented Gradients for Human Detection
Multiclass SVM and Applications in Object Classification
Image classification Given the bag-of-features representations of images from different classes, how do we learn a model for distinguishing them?
Handwritten digit recognition
Olivier Duchenne , Armand Joulin , Jean Ponce Willow Lab , ICCV2011.
Classification using intersection kernel SVMs is efficient Joint work with Subhransu Maji and Alex Berg Jitendra Malik UC Berkeley.
MIT CSAIL Vision interfaces Approximate Correspondences in High Dimensions Kristen Grauman* Trevor Darrell MIT CSAIL (*) UT Austin…
CS395: Visual Recognition Spatial Pyramid Matching Heath Vinicombe The University of Texas at Austin 21 st September 2012.
Computer Vision for Human-Computer InteractionResearch Group, Universität Karlsruhe (TH) cv:hci Dr. Edgar Seemann 1 Computer Vision: Histograms of Oriented.
Lecture 31: Modern object recognition
Activity Recognition Aneeq Zia. Agenda What is activity recognition Typical methods used for action recognition “Evaluation of local spatio-temporal features.
Computer Vision – Image Representation (Histograms)
Global spatial layout: spatial pyramid matching Spatial weighting the features Beyond bags of features: Adding spatial information.
Fast intersection kernel SVMs for Realtime Object Detection
More sliding window detection: Discriminative part-based models Many slides based on P. FelzenszwalbP. Felzenszwalb.
Discriminative and generative methods for bags of features
Computer Vision Group University of California Berkeley Shape Matching and Object Recognition using Shape Contexts Jitendra Malik U.C. Berkeley (joint.
Recognition using Regions CVPR Outline Introduction Overview of the Approach Experimental Results Conclusion.
Robust and large-scale alignment Image from
Poselets Michael Krainin CSE 590V Oct 18, Person Detection Dalal and Triggs ‘05 – Learn to classify pedestrians vs. background – HOG + linear SVM.
Image classification Given the bag-of-features representations of images from different classes, how do we learn a model for distinguishing them?
Beyond bags of features: Adding spatial information Many slides adapted from Fei-Fei Li, Rob Fergus, and Antonio Torralba.
Generic Object Detection using Feature Maps Oscar Danielsson Stefan Carlsson
CS294‐43: Visual Object and Activity Recognition Prof. Trevor Darrell Spring 2009 March 17 th, 2009.
Computer Vision Group University of California Berkeley Matching Shapes Serge Belongie *, Jitendra Malik and Jan Puzicha U.C. Berkeley * Present address:
A New Correspondence Algorithm Jitendra Malik Computer Science Division University of California, Berkeley Joint work with Serge Belongie, Jan Puzicha,
Matthew Brown University of British Columbia (prev.) Microsoft Research [ Collaborators: † Simon Winder, *Gang Hua, † Rick Szeliski † =MS Research, *=MS.
Generic object detection with deformable part-based models
Review: Intro to recognition Recognition tasks Machine learning approach: training, testing, generalization Example classifiers Nearest neighbor Linear.
Marcin Marszałek, Ivan Laptev, Cordelia Schmid Computer Vision and Pattern Recognition, CVPR Actions in Context.
Computational Photography Tamara Berg Features. Representing Images Keep all the pixels! Pros? Cons?
CSE 473/573 Computer Vision and Image Processing (CVIP) Ifeoma Nwogu Lecture 24 – Classifiers 1.
Handwritten digit recognition Jitendra Malik. Handwritten digit recognition (MNIST,USPS) LeCun’s Convolutional Neural Networks variations (0.8%, 0.6%
Svetlana Lazebnik, Cordelia Schmid, Jean Ponce
SVM-KNN Discriminative Nearest Neighbor Classification for Visual Category Recognition Hao Zhang, Alex Berg, Michael Maire, Jitendra Malik.
Lecture 31: Modern recognition CS4670 / 5670: Computer Vision Noah Snavely.
Pedestrian Detection and Localization
Classifiers Given a feature representation for images, how do we learn a model for distinguishing features from different classes? Zebra Non-zebra Decision.
Beyond Sliding Windows: Object Localization by Efficient Subwindow Search The best paper prize at CVPR 2008.
Evaluation of interest points and descriptors. Introduction Quantitative evaluation of interest point detectors –points / regions at the same relative.
Computer Vision Group University of California Berkeley On Visual Recognition Jitendra Malik UC Berkeley.
Handwritten digit recognition
Object detection, deep learning, and R-CNNs
Gang WangDerek HoiemDavid Forsyth. INTRODUCTION APROACH (implement detail) EXPERIMENTS CONCLUSION.
Histograms of Oriented Gradients for Human Detection(HOG)
Methods for classification and image representation
Locally Linear Support Vector Machines Ľubor Ladický Philip H.S. Torr.
Line Matching Jonghee Park GIST CV-Lab..  Lines –Fundamental feature in many computer vision fields 3D reconstruction, SLAM, motion estimation –Useful.
Presented by David Lee 3/20/2006
Goggle Gist on the Google Phone A Content-based image retrieval system for the Google phone Manu Viswanathan Chin-Kai Chang Ji Hyun Moon.
SUN Database: Large-scale Scene Recognition from Abbey to Zoo Jianxiong Xiao *James Haysy Krista A. Ehinger Aude Oliva Antonio Torralba Massachusetts Institute.
776 Computer Vision Jan-Michael Frahm Spring 2012.
More sliding window detection: Discriminative part-based models
Cascade for Fast Detection
CS262: Computer Vision Lect 09: SIFT Descriptors
Tues April 25 Kristen Grauman UT Austin
Presented by David Lee 3/20/2006
Learning Mid-Level Features For Recognition
Support Vector Machines and Kernels
Lit part of blue dress and shadowed part of white dress are the same color
Paper Presentation: Shape and Matching
Geometric Blur Descriptors for Point Correspondence
A Tutorial on HOG Human Detection
HOGgles Visualizing Object Detection Features
CS 1674: Intro to Computer Vision Scene Recognition
Edge detection f(x,y) viewed as a smooth function
A Graph-Matching Kernel for Object Categorization
Presentation transcript:

Digit Recognition using SVMS Jitendra Malik Lecture is based on Maji & Malik (2009)

Digit recognition using SVMs What feature vectors should we use? Pixel brightness values Orientation histograms What kernel should we use for the SVM? Linear Intersection kernel Polynomial Gaussian Radial Basis Function

Some popular kernels in computer vision x and y are two feature vectors

Kernelized SVMs slow to evaluate Decision function is where: Sum over all support vectors Kernel Evaluation Feature vector to evaluate Feature corresponding to a support vector l Arbitrary Kernel Histogram Intersection Kernel Cost: # Support Vectors x Cost of kernel computation

Complexity considerations Linear kernels are the fastest Intersection kernels are nearly as fast, using the “Fast Intersection Kernel” (Maji, Berg & Malik, 2008) Non-linear kernels such as the polynomial kernel or Gaussian radial basis functions are the slowest, because of the need to evaluate kernel products with each support vector. There could be thousands of support vectors!

Raw pixels do not make a good feature vector Each digit in the MNIST DATABASE of handwritten digits is a 28 x 28 pixel grey level image.

Error rates vs. the number of training examples

Technical details on orientation computation

Details of histogram computation

The 79 Errors

Some key references on orientation histograms D. Lowe, ICCV 1999, SIFT A. Oliva & A. Torralba, IJCV 2001, GIST A. Berg & J. Malik, CVPR 2001, Geometric Blur N. Dalal & B. Triggs, CVPR 2005, HOG S. Lazebnik, C. Schmid & J. Ponce, CVPR 2006, Spatial Pyramid Matching