Asaphus offers safety during autonomous driving, convenient and individualized driver-car interaction.
By recognizing who is using an appliance, Asaphus enables a convenient and individualized user experience.
Asaphus recognizes the driver and prevents theft. By alerting inattentive drivers, Asaphus improves road safety and helps to reduce insurance costs for commercial fleets.
The Asaphus Face Recognition Library can be embedded in ignition interlock devices; it makes regulatory compliance convenient and affordable.
At Asaphus Vision, we develop software that promotes the safety, convenience, and individuality of its users. We provide our customers with the best-performing and most innovative technology for face identification and gaze estimation on embedded devices. We are a university spin-off and have tight links to the research community.
Our mascot Asaphus kowalewskii is a trilobite that lived around 450 million years ago and features the prettiest eye stalks in the animal kingdom.
Offering a perfectly convenient and individual driving experience requires cars to recognize their driver. Driver identification allows the vehicle to adapt its settings to the driver, and to limit both vehicle performance and access to personal information for unknown drivers.
The Asaphus Face Recognition Library allows automotive suppliers to integrate driver identification functionality into their driver-monitoring systems. The Asaphus Face Recognition Library can be delivered for any development toolchain and operating system, and for single or dual near-infrared cameras. We optimize the library for specific hardware platforms and provide close support during product integration.
Implementing level 2 autonomous driving safely requires the vehicle to monitor the driver‘s alertness and ability to intervene at any time. For safe level 3 autonomous driving, vehicles are required to monitor the driver‘s attentional state and ability to intervene with limited lead time. Level 4 vehicles need to manage a safe handover from automomous to manual driving.
The Asaphus Face Recognition Library determines the driver‘s head pose, eye state, and eye gaze at high frame rates and with minimal requirements on computational resources. It provides basic building blocks for the most cost-effective, accurate, and highly-available driver-monitoring systems. We customize the library for any optical path, development toolchain, and computing hardware, and provide close support during product integration.
Facial identification provides a line of defense against the theft of delivery and vocational vehicles that is robust against possible negligent behavior of drivers. Driver distraction and drowsiness detection software can help fleet operators reduce the risk of accidents as well as insurance costs. The Asaphus Face Recognition Library can be embedded into existing telematics hardware. The software provides building blocks for the most cost-effective, accurate, and highly-available driver-monitoring and identification systems.
In order to offer the best possible user experience to each individual user, TV sets and other household appliances have to be aware who is using them. The Asaphus Face Recognition Library allows appliances to offer personalized recommendation, settings, and levels of user guidance. The Asaphus Face Recognition Library runs on a wide variety of embedded processors – such as an M4, A8, and A9 – and works under Linux, QNX, and any other operating system. It recognizes registered users and tracks their direction of gaze.
By verifying that a breath test is in fact taken by the registered user, facial recognition can improve road safety and eliminate the effort that today is spent on manual inspection of images. The Asaphus Face Recognition Library identifies users quickly on a wide variety of embedded processors – such as an M4, A8, and A9 – and works under Linux, QNX, and any other operating system. It determines calibrated identification probabilities and allows the system to adhere to defined false-positive rates. It can be integrated into existing systems that are equipped with a camera by way of a firmware update.
The Asaphus Face Recognition Library is a gaze-estimation, and face-identification software that is optimized for deployment in embedded systems. It offers an API for 3D head position and orientation, eye state, eye gaze, and identification.
The Asaphus Face Recognition Library is based on deep-learning technology; it is extremely efficient and fully self-contained. It runs at high frame rates on ARM A7, A8, A9 and other embedded processors with virtually any operating system. The software supports single and dual near-infrared cameras with VGA or higher resolution. It supports a wide range of camera mounting positions and is robust against varying head poses, bright daylight, and partial occlusions.
We provide our customers with the best-performing and most innovative technology for facial recognition on embedded devices. We are a spin-off company of the University of Potsdam, Germany, where we started our work on embedded face recognition in a research project in 2012.
Our team consists of experts in machine and deep learning, face recognition, and embedded software development.
Dr. Lenka Ivantysynova is the CEO of Asaphus Vision. She received her doctoral degree in business information systems in 2008 and her master’s degree in computer science in 2005 from Humboldt-Universität zu Berlin. During her doctoral studies, she conducted case studies in the automotive and manufacturing industries; she wrote her master’s thesis at Daimler Research (former DaimlerChrysler). As a business consultant, she consulted several large enterprises at the executive level on the integration of new technologies into their corporate IT strategies, and on major development projects.
Arvid Terzibaschian works on our core technology. Arvid is passionate about new computer vision and machine learning technologies and consantly explores ways to further improve our product. He has received his master’s degree (Dipl.-Inf.) in computer science from Humboldt-Universität zu Berlin. As a research associate at the University of Potsdam, he has worked on robust face recognition algorithms and has implemented the core technology of Asaphus Vision.
Dr. Peter Haider works on our core technology. Peter is passionate about solving mathematical puzzles. He applies his deep understanding of machine learning and computer vision with extreme determination and perseverance to overcome the most difficult technological hurdles. He received his master’s degree (Dipl.-Inf.) in computer science in 2006 from Humboldt-Universität zu Berlin and his PhD in machine learning in 2013 from the University of Potsdam.
Asaphus Vision cooperates closely with the machine learning group at the University of Potsdam.
Prof. Dr. Tobias Scheffer is a Professor of Computer Science at the University of Potsdam. He received his doctoral degree from Technische Universität Berlin in 1999. He has headed machine learning research groups at Humboldt-Universität zu Berlin, Germany, and the Max Planck Institute for Computer Science in Saarbrücken, Germany. His group’s research focuses on application-oriented machine-learning research. Together with Asaphus Vision, his group explores new algorithms for yet better-performing, more robust and efficient face recognition.
2015 we have won the Weconomy award which has been handed out to us by Franz Fehrenbach, chairman of the board of the Robert Bosch GmbH.
We have recieved the Exist government funding to finance our spinn of from the University of Potsdam, Germany.
2014 we have won the IKT Gründerwettbewerb. We are honored that the award was handed to us by Brigitte Zypries, parliamentary state secretary at the German federal ministry for economic afairs and energy.
Looking for your next big adventure? Are you passionate about solving complex and interesting problems? Join our team! Our team is fast and innovative, following agile methodologies and at the same time keeping the highest quality in all our processes.
If you are self-driven, passionate, a team-player and as enthusiastic about face recognition as we are please e-mail your application, including your CV and salary expectations, to jobs@asaphus.de.
Description:
As Software Engineer you develop innovative face recognition applications utilizing modern technologies in the area of machine learning and deep learning.
Qualifications:
Description:
As R&D Engineer you research, design, and implement new deep learning algorithms for your core product the Asaphus embedded Face Recognition Library.
Qualifications:
Description:
As intern you will be able to work on and evaluate our newest face recognition algorithms. You would work closely with our core researchers.
Qualifications:
Asaphus Vision GmbH
CEO: Dr. Lenka Ivantysynova
Fon: +49 30 850 191 77
Fax: +49 30 850 191 76
Email: contact@asaphus.de
Visiting Address:
Asaphus Vision
Bismarckstrasse 10-12
10625 Berlin | Germany
