High-Level Vision Face Detection CS332 Visual Processing Department

High-Level Vision Face Detection CS332 Visual Processing Department

High-Level Vision Face Detection CS332 Visual Processing Department of Computer Science Wellesley College Face detection: Viola & Jones Multiple view-based classifiers based on simple features that best discriminate faces vs. non-faces Most discriminating features learned from thousands of samples of face and non-face image windows Attentional mechanism:

cascade of increasingly discriminating classifiers improves performance 1-2 Viola & Jones use simple features Use simple rectangle features: I(x,y) in gray area I(x,y) in white area within 24 x 24 image sub-windows initially consider 160,000 potential features per sub-window!

features computed very efficiently Which features best distinguish face vs. non-face? Learn most discriminating features from thousands of samples of face and non-face image windows 1-3 Learning the best features weak classifier using one feature: x = image window

f = feature p = +1 or -1 = threshold (x1,w1,1) normalize weights (xn,wn,0) n training samples,

equal weights, known classes find next best weak classifier final classifier AdaBoost use classification errors to update weights

~ 200 features yields good results for monolithic classifier 1-4 Attentional cascade of increasingly discriminating classifiers Early classifiers use a few highly discriminating features, low threshold 1st classifier uses two features, removes 50% non-face windows later classifiers distinguish harder

Increases efficiency examples Allows use of many more features Cascade of 38 classifiers, using ~6000 features 1-5 Training with normalized faces 5000 faces many more non-face patches faces are normalized for

scale & rotation small variation in pose Viola & Jones results With additional diagonal features, classifiers were created to handle image rotations and profile views 1-7 Faces everywhere...

Recently Viewed Presentations

  • Chemical Bonding (short) - Chemistry Geek

    Chemical Bonding (short) - Chemistry Geek

    Use the Lewis structure to determine the geometry of the molecule. Electron arrangement establishes the bond angles. Molecule takes the shape of that portion of the electron arrangement. Charts look at the CENTRAL atom for all data! Think REGIONS OF...
  • 1.1 Introduction

    1.1 Introduction

    The equals sign (=) indicates that the variable should be initialized with the result of the expression to the right of the equals sign. The new keyword creates an object. Standard input object, System.in, enables applications to read bytes of...
  • Kitchen Tools - cfalls.org

    Kitchen Tools - cfalls.org

    Baking Pans / Casserole Dish . 9x13 and 8x8. To bake square cakes, desserts, or casseroles. Loaf Pan. ... Kitchen Shears. To open food packages or cut foods. Very sharp and stronger than scissors. Flatware. Utensils we eat with that...
  • Essential Concepts of Metabolism

    Essential Concepts of Metabolism

    ESSENTIAL CONCEPTS OF METABOLISM CHAPTER 5 Metabolism Anabolism Catabolism Energy Capture Oxidation Reduction Redox Metabolic Lifestyles — Autotrophy Photoautotroph Chemoautotroph Metabolic Lifestyles — Heterotrophy Photoheterotroph Chemoheterotroph Metabolic Lifestyles Summary Enzymes Role Structure Enzyme Sites Active Site Enzyme Inhibition Competitive Enzyme...
  • Update on the American Seed Trade Association &

    Update on the American Seed Trade Association &

    Breeding Methods: DNA markers, established heterotic pools, faster cycle time. Breeding Selection Criteria: Performance (yield, agronomics, disease) DUS Criteria: Morphology (non-performance visual observation) Standard morphological descriptors often fail to establish distinctness. PVP Office asks for additional evidence.
  • Présentation PowerPoint - WordPress.com

    Présentation PowerPoint - WordPress.com

    HOMOPHONES Réalisé par Lucie Gemme, Centre de Formation Générale La Croisée Pour le cours FRA-3048 Février 2009 Je vais a/à l'école a/à pied.. S'est/C'est en cherchant mon soulier qu'on/ont a/à troué mon chapeau. Ses/Ces chats son/sont disparus du berger blanc.
  • DO COMMUNITY-COLLEGE STUDENTS USE MORE CREDIBLE RESEARCH SOURCES

    DO COMMUNITY-COLLEGE STUDENTS USE MORE CREDIBLE RESEARCH SOURCES

    UC Blue Ash is the third-largest college (by enrollment) in the University of Cincinnati and the largest regional college campus in Ohio. Opened in 1967, the college is located in the northeastern corner of Hamilton County and serves residents of...
  • Distribution of Power

    Distribution of Power

    Explain how most countries have a mixed economy located on a continuum between pure market and pure command. c. Compare and contrast the basic types of economic systems found in Canada, Cuba, and Brazil. What is Economics? Economics is the...