Automated Face Detection Peter Brende David Black-Schaffer Veni Bourakov.

Slides:



Advertisements
Similar presentations
Applications of one-class classification
Advertisements

Pseudo-Relevance Feedback For Multimedia Retrieval By Rong Yan, Alexander G. and Rong Jin Mwangi S. Kariuki
Advanced Image Processing Student Seminar: Lipreading Method using color extraction method and eigenspace technique ( Yasuyuki Nakata and Moritoshi Ando.
Image Segmentation Image segmentation (segmentace obrazu) –division or separation of the image into segments (connected regions) of similar properties.
Foreground Modeling The Shape of Things that Came Nathan Jacobs Advisor: Robert Pless Computer Science Washington University in St. Louis.
Multiple People Detection and Tracking with Occlusion Presenter: Feifei Huo Supervisor: Dr. Emile A. Hendriks Dr. A. H. J. Stijn Oomes Information and.
fMRI data analysis at CCBI
ICIP 2000, Vancouver, Canada IVML, ECE, NTUA Face Detection: Is it only for Face Recognition?  A few years earlier  Face Detection Face Recognition 
Multiple testing Justin Chumbley Laboratory for Social and Neural Systems Research Institute for Empirical Research in Economics University of Zurich With.
Lecture 5 Template matching
Workshop on Earth Observation for Urban Planning and Management, 20 th November 2006, HK 1 Zhilin Li & Kourosh Khoshelham Dept of Land Surveying & Geo-Informatics.
Real-time Embedded Face Recognition for Smart Home Fei Zuo, Student Member, IEEE, Peter H. N. de With, Senior Member, IEEE.
3. Introduction to Digital Image Analysis
Automatic Face Recognition Using Color Based Segmentation and Intelligent Energy Detection Michael Padilla and Zihong Fan Group 16 EE368, Spring
 Image Search Engine Results now  Focus on GIS image registration  The Technique and its advantages  Internal working  Sample Results  Applicable.
Shadow Removal Seminar
Triangle-based approach to the detection of human face March 2001 PATTERN RECOGNITION Speaker Jing. AIP Lab.
Smart Traveller with Visual Translator for OCR and Face Recognition LYU0203 FYP.
[cvPONG] A 3-D Pong Game Controlled Using Computer Vision Techniques Quan Yu and Chris Wagner.
Face Processing System Presented by: Harvest Jang Group meeting Fall 2002.
1 Real Time, Online Detection of Abandoned Objects in Public Areas Proceedings of the 2006 IEEE International Conference on Robotics and Automation Authors.
Feature extraction Feature extraction involves finding features of the segmented image. Usually performed on a binary image produced from.
Brief overview of ideas In this introductory lecture I will show short explanations of basic image processing methods In next lectures we will go into.
Matthias Wimmer, Bernd Radig, Michael Beetz Chair for Image Understanding Computer Science Technische Universität München Adaptive.
Interactive Face Recognition (IFR) Nishanth Vincent Fairfield University Advisor: Professor Douglas A. Lyon, Ph.D.
Computer vision: models, learning and inference Chapter 6 Learning and Inference in Vision.
Tal Mor  Create an automatic system that given an image of a room and a color, will color the room walls  Maintaining the original texture.
06 - Boundary Models Overview Edge Tracking Active Contours Conclusion.
Information Extraction from Cricket Videos Syed Ahsan Ishtiaque Kumar Srijan.
EADS DS / SDC LTIS Page 1 7 th CNES/DLR Workshop on Information Extraction and Scene Understanding for Meter Resolution Image – 29/03/07 - Oberpfaffenhofen.
Overview Harris interest points Comparing interest points (SSD, ZNCC, SIFT) Scale & affine invariant interest points Evaluation and comparison of different.
Topic 10 - Image Analysis DIGITAL IMAGE PROCESSING Course 3624 Department of Physics and Astronomy Professor Bob Warwick.
Visual Inspection Product reliability is of maximum importance in most mass-production facilities.  100% inspection of all parts, subassemblies, and.
Joon Hyung Shim, Jinkyu Yang, and Inseong Kim
Phase Congruency Detects Corners and Edges Peter Kovesi School of Computer Science & Software Engineering The University of Western Australia.
ECE738 Advanced Image Processing Face Detection IEEE Trans. PAMI, July 1997.
A Statistical Method for 3D Object Detection Applied to Face and Cars CVPR 2000 Henry Schneiderman and Takeo Kanade Robotics Institute, Carnegie Mellon.
Digital Image Processing
Non-Photorealistic Rendering and Content- Based Image Retrieval Yuan-Hao Lai Pacific Graphics (2003)
Adaptive Rigid Multi-region Selection for 3D face recognition K. Chang, K. Bowyer, P. Flynn Paper presentation Kin-chung (Ryan) Wong 2006/7/27.
Veggie Vision: A Produce Recognition System R.M. Bolle J.H. Connell N. Haas R. Mohan G. Taubin IBM T.J. Watson Resarch Center Presented by Chris McClendon.
Eye regions localization Balázs Harangi – University of Debrecen Ciprian Pop – Technical University of Cluj-Napoca László Kovács – University of Debrecen.
Levels of Image Data Representation 4.2. Traditional Image Data Structures 4.3. Hierarchical Data Structures Chapter 4 – Data structures for.
Automatic License Plate Location Using Template Matching University of Wisconsin - Madison ECE 533 Image Processing Fall 2004 Project Kerry Widder.
Autonomous Robots Vision © Manfred Huber 2014.
Face Detection Using Skin Color and Gabor Wavelet Representation Information and Communication Theory Group Faculty of Information Technology and System.
GENDER AND AGE RECOGNITION FOR VIDEO ANALYTICS SOLUTION PRESENTED BY: SUBHASH REDDY JOLAPURAM.
Spatial Smoothing and Multiple Comparisons Correction for Dummies Alexa Morcom, Matthew Brett Acknowledgements.
Face Detection and Gender Recognition EE368 Project Report Michael Bax Chunlei Liu Ping Li 28 May 2003.
Wonjun Kim and Changick Kim, Member, IEEE
EE368 Digital Image Processing Face Detection Project By Gaurav Srivastava Siddharth Joshi.
Project Overview CSE 6367 – Computer Vision Vassilis Athitsos University of Texas at Arlington.
Face Detection Final Presentation Mark Lee Nic Phillips Paul Sowden Andy Tait 9 th May 2006.
Machine Vision Edge Detection Techniques ENT 273 Lecture 6 Hema C.R.
Face Detection Using Color Thresholding and Eigenimage Template Matching Diederik Marius Sumita Pennathur Klint Rose.
EE368: Digital Image Processing Bernd Girod Leahy, p.1/15 Face Detection on Similar Color Images Scott Leahy EE368, Stanford University May 30, 2003.
Semantic Alignment Spring 2009 Ben-Gurion University of the Negev.
Morphological Image Processing
Face Detection – EE368 Group 10 May 30, Face Detection EE 368 Group 10 Waqar Mohsin Noman Ahmed Chung-Tse Mar.
Face Detection In Color Images Wenmiao Lu Shaohua Sun Group 3 EE368 Project.
Over the recent years, computer vision has started to play a significant role in the Human Computer Interaction (HCI). With efficient object tracking.
Content Based Coding of Face Images
EE368 Final Project Spring 2003
EE368 Face Detection Project Angi Chau, Ezinne Oji, Jeff Walters 28 May, 2003.
Face Detection EE368 Final Project Group 14 Ping Hsin Lee
PRESENTED BY Yang Jiao Timo Ahonen, Matti Pietikainen
Scott Tan Boonping Lau Chun Hui Weng
Group 1: Gary Chern Paul Gurney Jared Starman
Counting Iron-Absorbed Small Intestinal Cells
Morphological Filters Applications and Extension Morphological Filters
Presentation transcript:

Automated Face Detection Peter Brende David Black-Schaffer Veni Bourakov

Primary Challenges Scale differences Scale differences Overlapping/obstructed faces Overlapping/obstructed faces Lighting variation Lighting variation

Implementation Overview 1.Color-based skin separation 2.Spatial analysis to generate candidate faces Morphological? Morphological? or Template Matching? or Template Matching? 3.Eliminate false hits Hands & arms, based on shape Hands & arms, based on shape Roof, based on texture Roof, based on texture Neck, based on relative position Neck, based on relative position

Color-based Skin Separation What color space to use? What color space to use? How to separate out the skin color? How to separate out the skin color?

Marginal Color Distributions

Parametric Separation Simple & Fast (h>0.98 or h 0.98 or h<0.01) Problems with non-linearity of HSV space in bright areas Problems with non-linearity of HSV space in bright areas

Full Joint-Probability Distribution We have enough data, so why not? We have enough data, so why not? Provides most accurate per-pixel classification. Provides most accurate per-pixel classification. Allows use to circumvent choosing a decision boundary. We can simply use Bayes rule. Allows use to circumvent choosing a decision boundary. We can simply use Bayes rule.

Slices from 3D Joint Distribution

Skin-probability Image Obtained From Applying Bayes Rule

Color-based Skin Separation What color space to use? What color space to use? HSV if separability of distributions is necessary HSV if separability of distributions is necessary How to separate out the skin color? How to separate out the skin color? Parametric is fast but loose in HSV Parametric is fast but loose in HSV Provides a binary mapping and requires choosing thresholds Provides a binary mapping and requires choosing thresholds Full PDF is accurate in any color space Full PDF is accurate in any color space Can be fast if done correctly (table lookup) Can be fast if done correctly (table lookup) No thresholds: produces a pure probability map No thresholds: produces a pure probability map

Spatial Analysis Method Morphological Morphological Obtain binary mask through thresholding Obtain binary mask through thresholding Perform morphological operations to separate and identify blobs corresponding to faces Perform morphological operations to separate and identify blobs corresponding to faces Difficult due to overlapping faces Difficult due to overlapping faces Template Match Template Match Search a scene for prototypical face image Search a scene for prototypical face image Need to decide which data to work with (luminance vs. skin probability) Need to decide which data to work with (luminance vs. skin probability)

Template Matching Using the skin-probability image: Greatly simplifies information content Greatly simplifies information content Simple information  simple algorithm Allows algorithm to focus on the single best facial clue: oval-shaped skin regions Allows algorithm to focus on the single best facial clue: oval-shaped skin regions Allows us to avoid creating a binary mask Allows us to avoid creating a binary mask

Correlation of simple template with Skin-probability image

Process of Inclusion/Elimination Iteratively pick ‘strong’ regions of the skin- probability image as faces: For each template search for matching face shapes (convolution peaks) For each template search for matching face shapes (convolution peaks) For each detected face, ‘subtract/erase’ the region from image to avoid duplicate detection For each detected face, ‘subtract/erase’ the region from image to avoid duplicate detection Stop when no significant skin regions remaining Stop when no significant skin regions remaining

Positive Detection and Elimination

Template and Threshold Selection Critical step of our algorithm Critical step of our algorithm Potential problems: Potential problems: Template matching face of a different size Template matching face of a different size Small template leads to double hits later Small template leads to double hits later Large template leads to missed faces Large template leads to missed faces Bad thresholds  low sensitivity or low specificity Bad thresholds  low sensitivity or low specificity Solution: Solution: Use templates of many sizes, going from largest to smallest Use templates of many sizes, going from largest to smallest Set threshold as high as possible without sacrificing sensitivity Set threshold as high as possible without sacrificing sensitivity

Algorithm Implementation 1. Load probability and template data 2. Down-sample the image by factor 2:1 3. Calculate the face-probability image by color 4. Remove hands/arms 5. Template match with skin-probability image 6. Eliminate false positive hits on necks of large faces 7. Remove patterned hits

Overall Results

Conclusions Recognition of face-shaped blobs from the skin- probability map works excellently Recognition of face-shaped blobs from the skin- probability map works excellently Requires that the skin colors be well known Requires that the skin colors be well known Requires that the general face sizes be well known Requires that the general face sizes be well known Our set of images was relatively consistent in terms of these factors Our set of images was relatively consistent in terms of these factors