Brodmann17, award-winning software company, will present a breakthrough with its deep-learning vision software solution that works at a fraction of current computing power needs – yet delivers the highest accuracy levels recorded for ADAS and autonomous driving vehicles.
Adi Pinhas, Brodmann17 CEO: “We provide a transformative DNN-based solution that can run computer vision even on low-power processors. Our patented technology achieves highest detection and recognition rates while requiring unprecedented few computing resources, making it an ideal solution to the most demanding challenges of autonomous vehicles and Advanced Driver Assistance Systems (ADAS).”
As part of their cooperation with Arm semiconductor, Brodmann17 have developed an ADAS system running on an Arm CPU A72, utilizing only a single core while maintaining high performance and excellent accuracy.
Adi Pinhas says “We bring truly life-changing technology to the automotive mainstream in an affordable way. Brodmann17 DNN does not require massive hardware changes, delivering Deep Neural Network solutions without compromising on quality or performance.”
Brodmann17 has partnered with global OEMs and Tier 1 leaders to define future automotive platforms and applications, such as vehicle and pedestrian detection, based on their innovative approach towards DNN and AI perception software.
Demand in the global ADAS market is fueled by applications for passenger car, regulation, and branding of government policy and rising demand for consumers’ safety.
Brodmann17 DNN software focus is on:
Core detection technology
Moving objects as: vehicles, pedestrians, cyclists
Static objects as: traffic lights, traffic signs
Free-space drivable area and lanes
Brodmann17 Supported ADAS functions include:
Forward Collision Warning (FCW)
Lane Departure Warning (LDW)
High Beam Control (HBC)
Pedestrian Collision Warning (PCW)
Cyclist Collision Warning (CCW)
Reverse Pedestrian Warning (RPW)
Efficient and safe autonomous vehicle systems require processing videos in high frame rates while maintaining perfect accuracy. Most common solutions today require massive GPU-based hardware and still struggle to achieve high performance without compromising accuracy.
Brodmann17’s pioneering technology transforms current systems into lightweight ones, ensuring perfect accuracy and performance while reducing the required compute power.
Adi Pinhas: “We have found the golden path that optimally balances performance, accuracy, hardware constraints and price, and can significantly improve the design of autonomous vehicle systems.”
TiE50 Awards: Brodmann17 is listed as a winner at the 2018 TiE50 Awards, Silicon valley’s premier annual award for the most successful technology startups.
Frost & Sullivan: 2017 Machine & Deep Learning for Computer vision Technology innovation Award
AutoSens: Finalists in the Software Innovation category for the 2018 AutoSens Awards.
Visit us at CES 2019, Israel pavilion Eureka Park Booth #51701