Out-of-plane Rotations Environment constraints ● Surveillance systems ● Car driver images ASM: ● Similarity does not remove 3D pose ● Multiple-view database.

Slides:



Advertisements
Similar presentations
Epipolar Geometry.
Advertisements

Overview Definition of Norms Low Rank Matrix Recovery Low Rank Approaches + Deformation Optimization Applications.
Geometry 2: A taste of projective geometry Introduction to Computer Vision Ronen Basri Weizmann Institute of Science.
Face Alignment with Part-Based Modeling
Two-View Geometry CS Sastry and Yang
1 pb.  camera model  calibration  separation (int/ext)  pose Don’t get lost! What are we doing? Projective geometry Numerical tools Uncalibrated cameras.
Automatic Feature Extraction for Multi-view 3D Face Recognition
University of Pennsylvania 1 GRASP CIS 580 Machine Perception Fall 2004 Jianbo Shi Object recognition.
IIIT Hyderabad Pose Invariant Palmprint Recognition Chhaya Methani and Anoop Namboodiri Centre for Visual Information Technology IIIT, Hyderabad, INDIA.
Xianfeng Gu, Yaling Wang, Tony Chan, Paul Thompson, Shing-Tung Yau
Two-view geometry.
Lecture 8: Stereo.
Image alignment Image from
View Morphing (Seitz & Dyer, SIGGRAPH 96)
Camera calibration and epipolar geometry
Image alignment Image from
Structure from motion.
Geometric Transformations CSE P 576 Larry Zitnick
Announcements Final Exam May 16 th, 8 am (not my idea). Practice quiz handout 5/8. Review session: think about good times. PS5: For challenge problems,
The 2D Projective Plane Points and Lines.
Structure from motion. Multiple-view geometry questions Scene geometry (structure): Given 2D point matches in two or more images, where are the corresponding.
Epipolar Geometry and the Fundamental Matrix F
Projective geometry Slides from Steve Seitz and Daniel DeMenthon
3D Hand Pose Estimation by Finding Appearance-Based Matches in a Large Database of Training Views
Multiple-view Reconstruction from Points and Lines
Single-view geometry Odilon Redon, Cyclops, 1914.
Projected image of a cube. Classical Calibration.
1 Invariant Local Feature for Object Recognition Presented by Wyman 2/05/2006.
May 2004Stereo1 Introduction to Computer Vision CS / ECE 181B Tuesday, May 11, 2004  Multiple view geometry and stereo  Handout #6 available (check with.
Lec 21: Fundamental Matrix
Camera parameters Extrinisic parameters define location and orientation of camera reference frame with respect to world frame Intrinsic parameters define.
Lecture 6: Feature matching and alignment CS4670: Computer Vision Noah Snavely.
Structure Computation. How to compute the position of a point in 3- space given its image in two views and the camera matrices of those two views Use.
PhD Thesis. Biometrics Science studying measurements and statistics of biological data Most relevant application: id. recognition 2.
CS 485/685 Computer Vision Face Recognition Using Principal Components Analysis (PCA) M. Turk, A. Pentland, "Eigenfaces for Recognition", Journal of Cognitive.
Multi-view geometry. Multi-view geometry problems Structure: Given projections of the same 3D point in two or more images, compute the 3D coordinates.
776 Computer Vision Jan-Michael Frahm, Enrique Dunn Spring 2013.
Multi-view geometry.
Geometry and Algebra of Multiple Views
Recognition and Matching based on local invariant features Cordelia Schmid INRIA, Grenoble David Lowe Univ. of British Columbia.
Homogeneous Coordinates (Projective Space) Let be a point in Euclidean space Change to homogeneous coordinates: Defined up to scale: Can go back to non-homogeneous.
Metrology 1.Perspective distortion. 2.Depth is lost.
Single View Geometry Course web page: vision.cis.udel.edu/cv April 9, 2003  Lecture 20.
Class 61 Multi-linear Systems and Invariant Theory in the Context of Computer Vision and Graphics Class 6: Duality and Shape Tensors CS329 Stanford University.
Geometric Transformations CSE 455 Ali Farhadi Many slides from Steve Seitz and Larry Zitnick.
A Flexible New Technique for Camera Calibration Zhengyou Zhang Sung Huh CSPS 643 Individual Presentation 1 February 25,
Two-view geometry. Epipolar Plane – plane containing baseline (1D family) Epipoles = intersections of baseline with image planes = projections of the.
Single-view geometry Odilon Redon, Cyclops, 1914.
776 Computer Vision Jan-Michael Frahm Spring 2012.
WAM “Writing About Math”
Optimal Features ASM Texture description based on Taylor series Grids centered at the landmarks for local analysis Non linear classifier (kNN) for inside-outside.
Image features and properties. Image content representation The simplest representation of an image pattern is to list image pixels, one after the other.
776 Computer Vision Jan-Michael Frahm Spring 2012.
Multi-view geometry. Multi-view geometry problems Structure: Given projections of the same 3D point in two or more images, compute the 3D coordinates.
3. Transformation
Lecture 07 13/12/2011 Shai Avidan הבהרה: החומר המחייב הוא החומר הנלמד בכיתה ולא זה המופיע / לא מופיע במצגת.
Parameter estimation class 5
Shape matching and object recognition using shape contexts
Geometric Transformations
Geometric Transformations
Multiple View Geometry for Robotics
Geometric Transformations
Geometric Transformations
Two-view geometry.
Multi-view geometry.
Homework: Study for Unit Test & Complete Test Prep Packet Learning Target: I can demonstrate how transformations and angle relationships impact geometric.
Recognition and Matching based on local invariant features
Geometric Transformations
Geometric Transformations
Presentation transcript:

Out-of-plane Rotations Environment constraints ● Surveillance systems ● Car driver images ASM: ● Similarity does not remove 3D pose ● Multiple-view database Other approaches ● Non-linear models ● 3D models: multiple views Database

Projective Geometry Geometric operations by means of linear algebra  2D points are 3- component vectors Multiple views of the same planar object can be related by homographies

Homographies Homographies hold both for object or camera movements The points must be coplanar H

Coplanar face model Silhouette points are excluded (out of main plane) Half the nose points are excluded (easy occlusion) First iteration: At least 8 correspondences to compute H (4 2D-points) Model Coordinates Image Coordinates

Image Matching ASM Image Model (Similarity) Gradient normal to the shape contour Projective transformations Do not preserve angles nor distance relationships H

Database (40 people)

Results PASM ASM Training and test on multi-view data Cross validation 7

Comparison to related work Ratios with respect to error on frontal images 8

Results training just a single view (frontal) Training set: Frontal Test set: Multilple views 9

Analysis of the single-view case

Conclusions on PASM If multi-view dataset available Almost invariant to rotations up to 60 degrees Training only on frontal views Considerably reduces (50%) variation of ASM due to viewpoint Left-right rotations better handled than up-down nodding Very difficult to compare to other results Points used for alignment can affect performance Not considerable for expected ASM precision 11

How reliable is the result?