Facefind Russian app Edit
- http://prtorrent.com/article/4988/russian+face+recognition+app+took+country+by+storm.html Searches a billion faces in one second.
Project oxford Edit
https://www.projectoxford.ai/sdk MSFT facial recognition.
http://bit.ly/181TWd Implement facial recognition software to determine if the driver behind the wheel is the owner or not. A car jacker's face will trigger an alert with his realtime picture delivered to the basestation via GPRS. The PC only records movement inside the car, should the hijacker ripout the camera the movement he made just before this will be sent to the basestation via GPRS. Nokia has implemented facial recognition in their phones to authenticate the user http://www.cellular-news.com/story/16331.php And this same technology can be used to authenticate the driver behind a wheel.
http://www.3vr.com sells facial recognition software for $1000 per camera.
Hitachi Kokusai Electric Edit
Surveillance system searches million of faces per second, looking for a match
Face.com, the company responsible for Facebook applications Photo Tagger and Photo Finder, lets you take any photo and quickly identify who is in it and where they are in the photo. This facial recognition is a boon to those tagging photos, and now Face.com is ready to bring a similar capability to the rest of the internet. May 3rd saw the launch of their new open API capable of scanning images and rapidly identifying the location, orientation, and identity of human faces
Driver authentication Edit
Latest advances Edit
The teams hope to give the first demonstration of FascinatE at the IBC expo in Amsterdam in September 2011, although the system will need further development before it can operate in real time.
Trade magazines Edit
Latest break through in facial recognition reduces the computing power and costs of facial authentication. http://www.gizmag.com/face-recognition-algorithm/12219/?utm_source=Gizmag+Subscribers&utm_campaign=79c31b9215-UA-2235360-4&utm_medium=email
commercial devices Edit
http://www.gizmag.com/bioscrypt-unveils-new-visionaccess-face-reader/9338/ Bioscrypt, a division of L-1 Identity Solutions has launched the next generation VisionAccess 3D face reader. Using the entire facial surface structure of a person’s face, the reader can authenticate users in less than one second, even in complete darkness.
Commercial software Edit
- http://briefcam.com/ Compressed thousands of hours from multiple cctv to identify the Boston bombers.
- http://fstbm.com/ Access control to apartments.
- http://www.aventuracctv.com/Intelligent-Video-Analytics-Software/ Aventura's Intelligent Video Analytics software has been designed to operate with both indoor and outdoor cameras, detecting a wide range of surveillance situations involving people, vehicles and other objects. A comprehensive range of detection filters provide the means to discriminate the object behaviors appropriate to each surveillance scene, meaning that classification is always accurate. Easy to set up, Aventura's Video Analytics Software automatically adjusts to the conditions it encounters, meaning that it is ideal for both experienced and untrained personnel to use, and can even be used within the home. This contrasts greatly with alternative software setups, which can have as many as 100 algorithm setup parameters just to adjust filters and adjust to lighting conditions. With factors such as this, it is easy to see that Aventura has been developed to produce real, working results, solely with the end-user in mind.
- http://avigilon.com/ videoIQ AI pattern recognition, used by Featherbrook estate S.A.
- http://www.pixlogic.com/ piXlogic provides visual search solutions that automatically analyze, index, and tag the contents of images and videos, providing an unparalleled level of search functionality to the Enterprise. The piXserve-ALERT software resides on its own server and interacts with other piXserve computer on the customer network. With piXserve‐ALERT, users can create a search query and convert it to an alert criteria. piXserve‐ALERT keeps track of what piXserve machines are indexing in near real time (broadcast video programs, security videos, web‐crawls, etc.) and when a match is made an e‐mail with a link is sent to the user. The user clicks on the link in the e‐mail and piXserve displays the image/video‐segment that has the content of interest. From the standard piXserve user interface, users can manage their alerts, share them with other members of their organization, review a history of triggers for each alert, create new ones, delete old ones, etc.
Open source Edit
* http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html * http://directory.google.com/Top/Computers/Security/Biometrics/Face_Recognition/
* http://www.dmoz.org/Computers/Security/Biometrics/Face_Recognition/ * http://www.facedetection.com/ * http://www.betaface.com/ * http://openbio.sourceforge.net/
* http://www.engineeringnews.co.za/article.php?a_id=57529 * Interoptic Networks founder Basia Nasiorowsk * http://www.tgdaily.com/content/view/39173/140/ Picaso from Google
AMD fusion of cpu and GPU Edit
http://www.theglobeandmail.com/news/technology/business-technology/amds-fusion-promises-big-jump-in-processing-power/article1938602/ For businesses or consumers, searching photos and video for specific images can be a time-consuming task. Whether it’s an auntie looking for shots of a favourite nephew or a journalist tracking down a clip of the celebrity of the day, it usually involves staring at the screen for long periods while going through file after file. Now facial recognition software can use the APU to quickly hunt through still and video images, frame by frame, to find a selected face.
Farmers bakkies Edit
Install the EmbeddedPc and everytime the bakkie's door gets opened a 10s delayed Jpeg image is gprs'd to the security company. Facial recognition software installed inside every bakkie would be ideal but probably to expensive for most. To reduce the cost of facial recognition software store the faces of hundreds of farmers on a central computer so that the software can verify their faces. Thus the security company would not need to manually verify the Jpeg image every time, which also has privacy implications - a computer does'nt care wether you combed your hair this morning or not.
This facial recognition only needs to be done everytime the door opens from a stationary position. It is impossible to hijack a vehicle driving at 60km/hour, it must be brought to a standstill. A Gps unit and electronic speed sensor will determine this.
Protecting your privacy Edit
Facial recognition inside a vehicle is an invasion of privacy if not implemented correctly. Secure the central authenticating PC in a room. The pc itself will have a camera trained on it so that anybody and everybody can make certain that no unauthorised access to the pc takes place. This enables us to make certain that any video footage stored on the central pc doesn't appear on http://www.youtube.com the next morning.
Checkup on your truckdriver Edit
Extremely small cameras or cameras embedded in the dashboard would allow an employer to check-up on his truck driver in realtime without the driver ever having to be aware of it. Facial recognition software is now so advanced that it can determine the wakeful state of the driver and sound an alarm if the driver were to fall a sleep.
Driver behind wheel Edit
Use facial recognition software such as http://www.polarrose.com to determine if the driver behind the wheel is the owner or not. A car jacker's face will trigger an alert with his realtime picture delivered to the basestation via GPRS. The PC only records movement inside the car, should the hijacker obstruct the camera the movement he made just before this will be sent to the basestation via GPRS. It's basically impossible to circumvent a facial recognition camera embedded in the dashboard. The software will immediatly try to authenticate the driver, any obstruction will result in a GPRS panic sent to the base station. http://www.extremecctv.com: ..."EB Series Facial Recognition Discreetly disguised as a light switch Complements wide-angle overview cameras for comprehensive surveillance. Download their document ExtremeCCTV-residential.pdf for details.
Community telephone exchanges Edit
Applications like FacialRecognition are expensive if it is installed at mutliple places. But by using a DsLam or other network and streaming a facial snapshot via a community telephone exchange across public roads to a central computer the cost of the software because almost nohting.
Store faces of people in street on database Edit
The PatrolVehicle captures the faces of anybody walking down the street should the CctvCameras not get a good view. Their faces are integrated with the SAPS database, Restaurant association and other communities setting up a TelephoneNetworkRollout. Should the same person frequent a street he will be flagged. If crook knows that his face is on some SAPS or other wanted database and we advertise that anybody walking down our streets faces will be captured, then such a would be burglar would rather go somewhere else.
Restaurant association Edit
A crook hitting a shop captured on the business owner PC must be uploaded in realtime to a national database along the lines of the restaurant initiative. www.restaurant.org.za RASA. They stream facial recognition via a satellite to a central database that analyses the pictures of their patrons. A camera mounted on a community telephone pole on the pavement can also link into the restaurant and SAPS database instantly identifying wanted suspects or ex-cons. Facial recognition on a single PC will enable automated tracking of people wandering aimlessly through quite streets. The computer will be able to plot his route, work out his estimated arrival time at the other end of the street compare it with a database of every persons face residing in a street.
* ImageProcessing * http://www.i-cube.co.za Number plate recognition * http://www.i-cubeinc.com/cameras.htm * http://www.polarrose.com/ Facial recognition in photos