Face recognition
Reconnaissance du visage
Face recognition uses mainly the following techniques :
- Facial geometry : uses geometrical characteristics of the face. May use several cameras to get better accuracy (2D, 3D...)
- Skin pattern recognition (Visual Skin Print)
- Facial thermogram : uses an infrared camera to map the face temperatures
- Vein : recognition of face vein pattern
- Smile : recognition of the wrinkle changes when smiling
- and some others, see further.
La reconnaissance du visage (ou reconnaissance faciale) utilisent principalement les techniques suivantes :
- La géométrie du visage : les caractéristiques géométriques de la face sont extraites. Peut réclamer plusieurs caméras pour obtenir une meilleure précision (pour faire du 3D par exemple).
- La structure fine de la surface de la peau (Visual Skin Print)
- Le thermogramme du visage : une caméra infrarouge permet de cartographier les températures
- Veines : reconnaissance de la forme des veines du visage
- Sourire : reconnaissance des variations des rides autours des lèvres lorsque l'on sourit
- et quelques autres techniques, voir plus loin.
And face recognition algorithms are facing the problem of makeup.
See also the specific page on smartphone using 3D face recognition.
For more information, as a start, get a look at Face Recognition Beyond the Visible Spectrum (paper also available on ResearchGate). You will find that it's possible to use the vascular network of the face, more détail in my page related to vein recognition.
- (2019 Nov) I recommend to read Bernard Smith's face recognition webpage.
Facial geometry
Géométrie du visage
Many different methods based on geometrical characteristics of the face have been developped such as "local feature analysis", "Eigenface or Principal Component Analysis", ...
Skin pattern recognition
Structure de la peau
Delean Vision (now Identix) developped Visual Skin Print, that uses pattern matching of the face skin.
Visual Skin Print™ relies on standard hardware - most web-cams and higher resolution mass-market video cameras, connected to a PC, will work. Visual Skin Print™ is based on a simple yet powerful idea : using the details of the skin for authentication.
Facial thermogram
Thermogramme du visage
Facial thermogram requires an (expensive) infrared camera to detect the facial heat patterns that are unique to every human being. Technology Recognition Systems worked on that subject in 1996-1999. Now disappeared.
Le thermogramme du visage requiert une caméra infrarouge (chère) pour détecter la répartition de température du visage particulière à chaque être humain. Technology Recognition Systems travailla sur ce thème vers 1996-1999, maintenant disparu.
"Face Recognition in Hyperspectral Images" (Zhihong Pan, Glenn Healey, Manish Prasad, Bruce Tromberg, 2003) is an article describing a variant using several wavelentghs.
- Some infrared face images
- Illumination Invariant Face Recognition Using Thermal Infrared Imagery (Solikinski & als).
- In 2002, Infrared Identification (IRID) got a NIST project using Mikron Infrared cameras.
Side effect : can detect lies
The image on the left shows his normal facial thermogram, and the image on the right shows the temperature changes when he lied.
L'image de gauche montre un thermogramme du visage sous conditions normales, et l'image de droite lors d'un mensonge, montrant les changements de température.
from Think thermally may 2002
Long-wave-infrared (LWIR) polarimetric-based thermal images
The polarimetric images show enhanced facial features, additional texture, and details that are not present in corresponding conventional thermal imagery.
- (2007) Polarimetric Thermal Imaging Fook Leong Loo Thesis
- (2014) Enhanced facial recognition for thermal imagery using polarimetric imaging Kristan P. Gurton & als
- (2015) Improving cross-modal face recognition using polarimetric imaging Nathaniel Short & als
- (2015) Changing the Paradigm in Human Identification
- (2016) A Polarimetric Thermal Database for Face Recognition Research Shuowen Hu & als
- (2018) Overview of polarimetric thermal imaging for biometrics Shuowen Hu & als
Vascular network, vein, blood vessels
réseau vasculaire
For more details, have a look at my page about vein recognition.
Smile recognition
Reconnaissance du sourire
The Stony Brook university system relies on probing the characteristic pattern of muscles beneath the skin of the face.
Guan takes two snaps of a person in quick succession, asking subjects to smile for the camera. He then uses a computer to analyse how the skin around the subject's mouth moves between the two images. The software does this by tracking changes in the position of tiny wrinkles in the skin, each just a fraction of a millimetre wide.
The data is used to produce an image of the face overlaid with tiny arrows that indicate how different areas of skin move during a smile. This movement is controlled by the pattern of muscles under the skin, and is not affected by the presence of make-up or the size of the subject's smile. The system is sensitive enough to produce such a map from muscle twitches even when people are trying to keep their expression unchanged.
Note : tested on 4 subjects (!).
L'université de Stony Brook met au point un système qui lit les plis de la peau provoqués par les muscles de la face lorsque l'on sourit.
Guan prend rapidement deux photos successives en demandant de sourire. Puis il analyse les mouvements de la peau autours de la bouche, de l'ordre du mm.
L'image du visage est recouvert par de petites flèches qui indiquent les mouvements relevés en souriant. Ces mouvements sont controlés par les muscles sous la peau, peu susceptibles d'être affectés par le maquillage ou l'amplitude du sourire. Le système serait suffisamment sensible pour mesurer les changements même si le sujet essaye de garder son expression inchangée.
Nota : test réalisé sur 4 sujets (!).
- Nature article (March '04)
Dynamic facial features
Reconnaissance faciale dynamique
In "A Dynamic Approach to Face Recognition" paper, a new method for face recognition is proposed, which is based on dynamic instead of static facial features. They track the motion of certain features on the face during a facial expression and obtain a vector field that characterizes the deformation of the face.
In the training process, two high-resolution images of an individual, one with a neutral expression and the other with a facial expression, like a subtle smile, are taken to obtain the deformation field.
This is a Joint Project between Stony Brook University (SUNY) and New York State Strategic Center for Port and Maritime Security, SUNY Maritime.
Il s'agit de détecter les déformations du visage liées à une expression du visage.
3D face recognition with LIDAR
- SVI StereoVision Imaging, Inc..
Product : The CES-S400 is the world’s longest-range commercial biometric system, capturing 2D, 3D and 4D facial data to identify targets at standoff distances. Our patented Dual Chirp FMCW 4D LiDAR provides unprecedented tactical ranges up to 30 meters.
brochure (2020) - (2021 Jan) SVI Teams with IDEMIA NSS to Offer World’s Most Advanced FMCW 4D LiDAR Biometric Recognition Solution.
Makeup challenges
Makeup challenges automated face recognition systems
How To Hide Your Face From Big Brother? Try Sunglasses.
More information
The following information is becoming out of date, it is just kept here "for the records". It is likely that most links are broken...
Labs :
- Mitsubishi MERL
- MIT face recognition
- Person Identification via Face Recognition MIT
- CSU Colorado State University (with source code)
- IBM face recognition
Tests :
Commercial products
- 3D Biometrics
- A4Vision 3D facial reconstruction
- Acsys VeraGuard, VeraShield...
- Attrasoft
- Authenmetric offers an infrared face recognition system
- Aurora Computer Systems
- Biometrica Systems
- Biovisec
- Brillertek
- Cognitec FaceVACS, SmartGate
- CyberExtruder 3D morphable model, Mugshot Pro
- C-VIS FaceSnap, FaceCheck...
- Dicut
- Dream MIRH
- Geometrix FaceVision
- HumanScan
- Identix (merged with Visionics) FaceIt, ABIS... (10-MAR-2004 : acquired Delean Vision VSP Visual Skin Print)
- Image Metrics Face DB based on Optasia
- Imagis Technologies ID-2000
- ImageWare Systems EPI builder, EPI suite...
- Infrared Identification (IRID)
- Neven Vision (formerly Eyematic)
- Neurodynamics Tridentity
- Neurotechnology VeriLook
- Nippon-Rad
- Opto-biometrics (Canada, created 2004)
- QiSQi Silent Watch
- Titanium Face-OK, TI-Face
- Viisage has acquired ZN. FacePASS, FaceEXPLORER, FaceFINDER...
- VisionSphere Technologies FaceCam, It's me...
Applications
- (2008 Sep) Google announced the latest version of its Picasa photo album software which uses facial recognition to track who is in each photo... via Blorge
Built-in devices
- (2006 Jan) Thakral dit offers the Snappy, a USB face recogniser. In fact a 1.3 megapixel USB camera.
- (2006 Jul) FTEC offers a full range of notebooks with face recognition.
- (2006 Jul) Founder S620 notebook with face recognition
- (2006 Dec) The Lenovo Y300 & Y500 notebooks use the integrated webcam for face recognition (by Veriface).
- (2007 Mar) RCG proposes the SecureBook Fx9 and the SecureBook Fx18 with face recognition using the 1.3 megapixel built-in camera.
- (2007 Sep) NEC proposes the LaVie L & C notebooks with "Face Pass".
- (2008 Jan) The Asus U2E notebook offers the Asus Smart Logon Face Recognition Technology.
- (2009 Mar) Lenovo Y650 with Veriface
- (2015 Jul) While Windows Hello supports fingerprint readers to log into Windows 10, the key new feature is face recognition. Using special infrared cameras, Windows Hello will scan a face and allow you to sign into a Windows 10 machine. Regular webcams won’t work with the feature, and you’ll need a machine that ships with Intel’s new RealSense 3D camera.
- (2017 Sep) Apple introduces Face ID with its iPhone X.