The human face plays an important role in our social interaction, conveying people’s identity. Using the human face as a key to security, biometric face recognition technology has received significant attention in the past several years due to its potential for a wide variety of applications in both law enforcement and non-law enforcement.
As compared with other biometrics systems using fingerprint/palmprint and iris, face recognition has distinct advantages because of its non-contact process. Face images can be captured from a distance without touching the person being identified, and the identification does not require interacting with the person. In addition, face recognition serves the crime deterrent purpose because face images that have been recorded and archived can later help identify a person.
NEC’s face recognition technology can be implemented as a functionally independent application, or seamlessly integrated into new or existing biometric security solutions by system integrators and solution providers.
Fast & Accurate Face Recognition
GLVQ based multiple-matching face detection
Combination of eye-zone extraction and facial recognition
Recognition based on neural network technology
Short processing time, high recognition rate
Recognition regardless of vantage point and facial changes (glasses, beard, and expression)
Optimal results through Adaptive Regional Blend Matching (ARBM) technology
Extraction of similar facial areas
Identification and authentication based on individual facial features
Easy adaptation to existing IT systems
Flexible integration into many types of video monitoring systems
Simple connection to NEC AFIS
Supporting diverse graphic and video formats as well as live cameras
Diverse Application Areas
NEC’s biometrics face recognition process has a highly diverse range of applications, extending from crime-fighting, border control, to access control for sensitive areas.
NEC's Face Recognition Technology Ranked No. 1 in NIST Testing for the fourth consecutive times
NEC's face recognition technology achieved the highest performance evaluation in the “Face in Video Evaluation (FIVE, *1) 2017” testing performed by the U.S. National Institute of Standards and Technology (NIST) (*2). Results were released in NIST's Interagency Report 8173: Face In Video Evaluation (FIVE) Face Recognition of Non-Cooperative Subjects.
NEC's face recognition technology took first place for the fourth consecutive time following the 2009 Multiple Biometric Grand Challenge (MBGC 2009), 2010-2011 Multiple Biometrics Evaluation (MBE 2010-2011), and 2013 Face Recognition Vendor Test (FRVT 2013).
Video face recognition technology identifies the faces of moving subjects in real-time as they walk naturally without stopping in front of a camera. The benefits of high-speed video analysis enabled by this technology include the prevention of potential incidents through detection of suspicious individuals and recognition of individuals at the gateways of critical facilities.
Using video images from standard cameras for face recognition requires highly-advanced techniques when compared to still images. This is because images are greatly influenced by environmental conditions, such as camera location, image quality, lighting and subject size, in addition to the behavior of a subject, including walking speed, face direction and sight line.
To achieve reliable face recognition of a video image, NEC developed feature point extraction technology that enables enhanced face recognition to a level where an individual can be identified with high precision from within a group, even if their face is partially hidden, or the image is taken from different angles. NEC's face recognition technology also uses deep learning technologies for face matching to increase accuracy to a level where an individual can be identified by a low resolution face image captured by a distant camera.
"It is a great honor that our face recognition technology received a first place evaluation for the fourth consecutive time from the NIST," said Masakazu Yamashina, Senior Vice President, NEC Corporation. "We will further expand global business by offering innovative biometric solutions, such as the video face recognition system, throughout a wide range of fields, including security, transportation, finance and retail."
Evaluation examples from the NIST FIVE testing
Entry-exit management at an airport passenger gate
Tests were conducted to recognize one individual at a time as they walk through an area without stopping or acknowledging the camera. NEC's face recognition technology won first place with a matching accuracy of 99.2%. The error rate of 0.8% is less than one-fourth of the second place error rate.
Detection of suspicious individuals at an indoor stadium
Tests were conducted with an individual situated far from the camera with their face direction changing frequently. NEC's face recognition technology won first place with an error rate half that of the second place error rate.
For nearly 30 years, NEC has been developing face recognition technology by defining it as one of the key technologies to help achieve a safer and more secure society. The technology has now been implemented in more than 100 systems in 40 countries worldwide. NEC will continue to develop solutions that leverage face recognition technology as part of its portfolio of AI technologies "NEC the Wise" (*3), and to deliver the solutions for a wide range of fields, including social infrastructure and private sector facilities.
NEC ranked No.1 in accuracy in both ideal and difficult environmental conditions for face recognition at Face in Video Evaluation 2017
- (*2)Results shown from the Face In Video Evaluation (FIVE), the Multiple Biometric Grand Challenge (MBGC), the Multiple Biometric Evaluation (MBE) and the Face Recognition Vendor Test (FRVT) do not constitute endorsement of any particular product by the U.S. Government.
Generalized Matching Face Detection Method (GMFD)
NEC’s face recognition technology utilizes the GMFD method that provides high speed and high accuracy for facial detection and facial features extraction. The main logic for facial recognition within GMFD is a modified Generalized Learning Vector Quantization (GLVQ) algorithm, which searches and selects face area candidates after the generation of potential eye pairs. GLVQ is based on a neural network and is not easily fooled by attempts to conceal identity via the usage of caps, hats, sunglasses, etc.
Perturbation Space Method (PSM)
NEC has developed the PSM algorithm that converts two-dimensional images (e.g., photographs) into three-dimensions (such a process is called “Morphing”). The three-dimensional representations of the head are then rotated in both the left-to-right and up-and-down directions. Further processing applies differing illumination across the face, which greatly enhanced the chances of a query “faceprint” for matching against its true mate from the database.
Adaptive Regional Blend Matching (ARBM) Method
Thanks to the PSM algorithm, the general range of facial poses and illumination has ceased to present major problems. However, the range of variation of different facial parts is still a challenge.
To reduce the impact of adverse local changes (e.g., varying facial expression caused by smiling and blinking eyes, and intentional changes caused by the wearing of caps, hats and glasses), NEC’s face recognition technology utilizes the ABRM algorithm, which reduces the impact of such local changes during the matching process. The minimization of the local changes guarantees the overall face recognition accuracy.