Last update : December 15, 2014
Face Recognition Definitions
A face recognition system is a computer application for automatically identifying or verifying a person from a digital image or from a video frame. One of the ways to do this is by comparing selected facial features (landmarks) from the image with a facial database.
Face recognition is used in biometrics, video surveillance, human computer interfaces and image database management (digital photo albums).
The prior step to face recognition is the accurate detection of human faces in arbitrary scenes. Face detection is the related computer technology that determines the locations and sizes of human faces in digital images. This is an effortless task for humans, but requires a great effort for computers.
Face detection is followed by normalization and extraction which leads to the final recognition. A next step could be the interpretation of the face. This understanding of an individual’s face is called face perception. The proportions and expressions of the human face are important to identify origin, emotions, health qualities, social information and accessories (glasses, beards, …).
Face detection is a general case of face localization and a specific case of object-class detection. In object-class detection, the task is to find the locations and sizes of all objects in an image that belong to a given class. Some digital cameras use face detection for autofocusor to take automatically a second picture if someone closed his eyes during exposure. Today face detection is a standard feature in various image edition programs, for instance in Adobe Photoshop Elements, Synology Photostation, Apple iPhoto, …
Face Recognition Promotion
In the years 2000, 2002, 2006, 2012 and 2013, the US National Institute of Standards and Technology (NIST), an agency of the U.S. Department of Commerce, organized a Face Vendor Recognition Test (FRVT). These tests provided independent government evaluations of commercially available and prototype face recognition technologies. These evaluations are designed to provide U.S. Government and law enforcement agencies with information to assist them in determining where and how facial recognition technology can best be deployed. In addition, FRVT results help identify future research directions for the face recognition community.
The US government conducted two other projects to promote and advance face recognition technologies. The FERET program started in September 1993 and ended in 1997. The Face Recognition Grand Challenge (FRGC) ran from May 2004 to March 2006. The test databases and results of FRVT and FRGC are still available for researchers.
There are other actors who promote face recognition technologies. The IEEE conference on Automatic Face and Gesture Recognition is the premier international forum for research in image and video- based face, gesture, and body movement recognition. Its broad scope includes advances in fundamental computer vision, pattern recognition, computer graphics, and machine learning techniques relevant to face, gesture, and body action, new algorithms, and analysis of specific applications. The Eleventh conference (FG 2015) will be held in Ljubljana, Slovenia in may 2015.
Another notable event in the domain of face recognition was the takeover in June 2012 by Facebook of Face.com Inc, an Israel-based technology company that developed a platform for efficient and accurate facial recognition in photos uploaded via web and mobile applications. One month later, the face.com API was shut down, which left some 45.000 developers stranded, and saw the developer community demanding an open source alternative. Six years earlier, Google acquired Neven Vision, whose technology for face recognition was integrated in Picasa Web Albums in September 2008.
Face Recognition Technologies
Face recognition is based on three approaches :
- geometric : this approach is based on geometrical relationship between facial landmarks (distances and angles between eyes, nose, mouth , eyebrows, chin, jaw, cheekbones, …).
- photometric : this is a statistical approach that distills an image into values and compares the values with templates to eliminate variance.
- 3D : three-dimensional (3D) face recognition is based on distinctive features such as curves of eye sockets, chin, nose, tissues and cheekbones. This approach allows to identify a face from a range of viewing angles, including a profile view.
Popular recognition algorithms include :
- Active Appearance Model (AAM)
- Bayesian Framework
- Boosting & Ensemble Solutions
- Elastic Bunch Graph Matching (EBGM)
- Evolutionary Pursuit (EP)
- Hidden Markov Models (HMM)
- Independent Component Analysis (ICA)
- Linear discriminant analysis (LDA)
- Kernel Methods (KM)
- Principal Component Analysis (PCA)
- Trace Transform
- Support Vector Machine (SVM)
Web based solutions
Animetrics : FaceR Identity Management System
Animetrics develops next-generation 2D-to-3D face recognition, identity management and biometrically enabled search engine products for authentication and search which links photographic clouds, cameras, video and image stores via Web-service based facial search engines. Animetrics is poised for strong growth in the Government market for mobile and intelligence, and the Commercial & Consumer markets for mobile, social networking and security by revolutionizing human interfaces to computing devices and social media.
APICloud.Me : FaceRect and FaceMark
FaceRect is a powerful and completely free API for face detection. FaceMark is a powerful API for facial feature detection. It finds 68 points for a frontal face and 35 for a profile one. FaceMark detects landmarks for faces on the image specified by URL or uploaded as a file and produces JSON output containing a vector of facial landmarks and orientation for each face found.
Betaface : Face Detection & Recognition
You can communicate with the Betaface webservice from any platform, any OS that will allow you to send and receive HTTP messages. The Website is owned and operated by Software Oleksandr Kazakov, Munich, Germany.
BioID : be recognized
BioID says to be the world’s first webcam-based personal recognition system for web and mobile users : convenient, reliable, secure. BioID provides a playground as test bed for the BioID Web Service, a collection of simple API calls enhancing existing systems with powerful biometrics (face recognition and voice recognition). You can sign up for a free My BioID account. Dr. Robert Frischholz, the author of the Face Detection Homepage, is the Chief Technical Officer (CTO) of BioID.
BiometryCloud : Face recognition for developers
BiometryCloud provides frameworks and plugins that, combined with their pattern recognition API’s, result in the most simple, fast and robust identity platform on the market. The service will be launched soon, you can subscribe to a private beta now.
HP Lab’s Multimedia Analytics Platform : beta
HP offers the Multimedia Analytics Platform Site as a place where you can experiment with some of the latest technologies that are under development at HP Labs, using your own media or sample pictures they provide. The following services are available : Face Detection, Face Verification, Face Demographic, Feature Extraction, Image Matching and Image Collage.
Lambda Labs : Face Recognition API βeta
Lambda Labs, Inc. was founded in the spring of 2012 in San Francisco and is now looking for the best hackers, academics, hustlers, and financiers to join them. A freemium subscription is offered to test the API which provides face recognition, facial detection, eye position, nose position, mouth position, and gender classification.
ReKognition : Integrated visual recognition solution based in Cloud
ReKognition is an open API platform provided by Orbeus, which helps developers easily incorporate the state-of-the-art computer vision capabilities of facial recognition, object recognition and scene understanding into your app or software. Orbeus is a computer vision company that empowers computers to see better than humans. Orbeus says to be the first and only company to provide a proprietary cloud-based image analysis solution that makes sense of faces, scenes and objects all together. You can sign up for a free Play Plan. An iOS beta app was available.
Sky Biometry : Cloud-based Face Detection and Recognition API
SkyBiometry introduces an cloud-based Face Detection and Recognition API. At the core of the product is one of the worlds best algorithms for the detection and recognition of faces that is successfully used in many commercial products around the globe. A free subscription is available.
More web services
Cybula FaceEnforce : 3D Face Recognition system
Cybula’s FaceEnforce system represents the leading edge technology for 3D facial biometrics.
FaceFirst : The Leader in Facial Recognition Technology
FaceFirst uses the world’s most highly developed facial recognition software. It is patent-pending technology rated number one by the 2006 Face Recognition Vendor Test conducted by the U.S. Commerce Department’s National Institute of Standards and Technology.
Luxand : FaceSDK 5.0
Luxand Inc. is a private hi-tech company formed in 2005. The company provides a broad range of facial feature recognition solutions. The company’s products and technologies are used by biometric identification and security companies, banks, the entertainment industry, medical and cosmetic industry, at online entertainment portals, chat rooms, and movie Web sites around the globe.
VeriLook facial identification technology is designed for biometric systems developers and integrators. The technology assures system performance and reliability with live face detection, simultaneous multiple face recognition and fast face matching in 1-to-1 and 1-to-many modes.
RoboRealm 2.67.45 : vision for machines
RoboRealm is an application for use in computer vision, image analysis, and robotic vision systems. RoboRealm’s face detection module is used to detect the presence of faces within an image.
Sensible Vision FastAccess SDK
FastAccess is facial recognition software for secure access to computing devices which reduces the need to manually log in to the device while maintaining strong security.
Free open source solutions
CCV 0.6 : A Modern Computer Vision Library
One core concept of ccv development is application driven. As a result, ccv end up implementing a handful state-of-art algorithms. It includes a very fast detection algorithm for rigid object (face etc.), an accurate object detection algorithm for somewhat difficult object (pedestrian, cat etc.), a state-of-art text detection algorithm, a long term object tracking algorithm, and the long-standing feature point detection algorithm.
The FaceRecLib is an open source platform that is designed to run comparable and reproducible face recognition experiments. This library is developed at the Biometrics group at the Idiap Research Institute. The FaceRecLib is a satellite package of the free signal processing and machine learning library Bob (version 2.0.0), and some of its algorithms rely on the CSU Face Recognition Resources.
OpenCV 3.0 : Open Source Computer Vision
OpenCV is released under a BSD license and hence it’s free for both academic and commercial use. It has C++, C, Python and Java interfaces and supports Windows, Linux, Mac OS, iOS and Android. OpenCV was designed for computational efficiency and with a strong focus on real-time applications. Written in optimized C/C++, the library can take advantage of multi-core processing.
OpenBR 0.5.0 : Open Source Biometric Recognition
OpenBR is a framework for investigating new modalities, improving existing algorithms, interfacing with commercial systems, measuring recognition performance, and deploying automated biometric systems. The project is designed to facilitate rapid algorithm prototyping, and features a mature core framework, flexible plugin system, and support for open and closed source development. Off-the-shelf algorithms are also available for specific modalities including Face Recognition, Age Estimation, and Gender Estimation. OpenBR requires OpenCV and Qt.
The following liist provides links to websites with additional informations about face recognition :
- The Face Detection Homepage, by Robert Frischholz
- The Face Recognition Homepage, by Mislav Grgic and Kresimir Delac
- List of 50+ Face Detection / Recognition API’s, libraries and software, by Mashape
- Detecting Faces in Images : A Survey, by Ming-Hsuan Yang
- Neural Networks for Face Recognition, by Tom Mitchell (1997)
- Reconnaissance des émotions et des expressions faciales, par Damien Leroux
- CSU Face Recognition Homepage, by Colorado State University