New version page

CORNELL CS 6670 - Lecture 18: Recent work in recognition

Upgrade to remove ads
Upgrade to remove ads
Unformatted text preview:

Lecture 18: Recent work in recognition (slides adapted from Dan Huttenlocher, Mark Everingham, and James Hays)AnnouncementsObject recognitionSlide 4Slide 5Slide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Slide 16Slide 17Slide 19Slide 21Slide 22Slide 23Slide 24Slide 25Slide 26Slide 27Slide 28Best localization methodsHoG featuresFlexible Spatial Template (UofC-TTI)Six-component car modelSix-component person modelGIST descriptorsGIST DescriptorSlide 36Slide 37Slide 38Slide 39Slide 40Slide 41Slide 42Scene Completion AlgorithmSlide 44Slide 45Slide 46Slide 47Slide 48Slide 49Slide 50Slide 51Slide 52Slide 53Slide 54Slide 55Slide 56Projective geometryLecture 18: Recent work in recognition (slides adapted from Dan Huttenlocher, Mark Everingham, and James Hays)CS6670: Computer VisionNoah SnavelyAnnouncements•Project 3: Eigenfaces–due Wednesday, November 11 at 11:59pmObject recognition•Category recognition has been the focus of extensive research in the past decade•Extensive use and development of machine learning techniques, better features•Moderate-scale datasets derived from the Web–PASCAL VOC: 20 object categories, > 10K images, > 25K instances, hand-labeled ground truth, annual competitions•Twenty object categories (aeroplane to TV/monitor) •Three challenges:–Classification challenge (is there an X in this image?)–Detection challenge (draw a box around every X)–Segmentation challengeisis there a cat?Chance essentially 0Best localization methods•Sliding window-style classifiers –SVM, Adaboost–Flexible spatial template: “star model”•Separate classifiers by viewpoint•Use of context in classifiers•Local features–HoG, SIFT, local histograms of gradient orientationsHoG features•Image partitioned into 8x8 blocks•In each block, compute histogram of gradient orientationsFlexible Spatial Template (UofC-TTI)•Hierarchical model [Felzenszwalb et al 2008]–Coarse template for finding the root part–Fine-scale templates connected by springs–Learning automatically from labeled bounding boxes•Separate models per viewpointSix-component car modelroot filters (coarse)part filters (fine) deformation modelsside viewfrontal viewSix-component person modelGIST descriptors•Global descriptor computed from entire image [Oliva and Torralba 2001]GIST DescriptorHays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Criminisi, Perez, and Toyama. Region filling and object removal by exemplar-based inpainting. IEEE Transactions on Image Processing. 2004.Hays and Efros, SIGGRAPH 2007Criminisi et al. resultHays and Efros, SIGGRAPH 2007Criminisi et al. resultHays and Efros, SIGGRAPH 2007Scene Completion AlgorithmInput image Scene DescriptorImage Collection200 matches20 completionsContext matching+ blending……Hays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Hays and Efros, SIGGRAPH 2007Projective geometryReadings•Mundy, J.L. and Zisserman, A., Geometric Invariance in Computer Vision, Appendix: Projective Geometry for Machine Vision, MIT Press, Cambridge, MA, 1992, (read23.1-23.5,23.10)–available online: http://www.cs.cmu.edu/~ph/869/papers/zisser-mundy.pdfAmes


View Full Document
Download Lecture 18: Recent work in recognition
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Lecture 18: Recent work in recognition and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Lecture 18: Recent work in recognition 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?