Automotive Vision Based On Deep Learning Technology To Elevate Advanced Driver-Assistance System

James Dargan

Area 17

Did you know the brain is comprised of 52 areas and grouped into 11 histological zones? Yes, originally defined and numbered by German neurologist Korbinian Brodmann, who mapped out the cerebral cortex, these have become known as Brodmann areas.

Brodmann employed a variety of criteria to map out the human brain, with particular attention focused on both gross anatomical features and cortical micro-structures.

In Brodmann’s theory, he suggested that areas with different structures performed different functions. Later, some of these areas were correlated to nervous functions, as in the following examples:

• *Brodmann area 41 and 42 in the temporal lobe, related to hearing

• Brodmann area 45 and 44 overlap with the Broca’s area for language in humans

• Brodmann area 1, 2, and 3 in the postcentral gyrus of the parietal lobe (the somatosensory region)

• Brodmann area 4 in the precentral gyrus of the frontal lobe (the primary motor area)

• Brodmann area 17 and 18 in the occipital lobe (the primary visual areas).*

Area 17, whether pure pseudo-science or rather based on rigorous scientific experimentation and observation, is interesting as it has been the inspiration of one Israeli deep learning (DL) startup based in Tel Aviv.

Embedded deep learning perception software designed for the automotive market

 — Brodmann17

Brodmann17

Brodmann17 was founded in 2016 by Adi Pinhas, Amir Alush and Assaf Mushinsky. Its DL and computer-vision algorithms — designed on AI that is revolutionizing safety in mobility — were built from the ground up utilizing patented DL technology designed for automotive vision and is reported to save an incredible 95% computer power at the same time.

Brodmann17 achieves this with engineered core deep-learning algorithms and patented new neural-network technology, based on weights and calculation sharing, leveraging this technology to reduce the number of unique calculations within the NN so that a very large number of recurring calculations may be shared between neurons, creating a state-of-the-art Advanced driver-assistance system (ADAS).

The trump card of this leads to

• Extended runtime and power

• Software only solution for any processor

• Cutting-edge accuracy even when using low power processors

“At Brodmann17, we are building on those advancements [deep learning technology], as well as developing our own algorithms and methods to bring the state of artificial intelligence to the automotive industry. And this is what Brodmann17 is working on.”

 — Dr. Amir Alush, CTO & CO-Founder, Brodmann17

Employing such technology is the range of products Brodmann17 provides, which includes

ADAS for fleet management, the Aftermarket ADAS Software Suite, the Front Active Safety Software Suite, and the Rear/Surround Camera ADAS Software.

Adi Pinhas is the first of three Co-founders of Brodmann17 who has founded two startups, one in intelligent IP video surveillance and security solutions and the other in visual-search and image-recognition, prior to starting Broadmann17. He has an M.Sc. in Computer Vision from Tel Aviv University.

An expert in DL and computer vision, Amir Alush is Broadmann’s CTO and the second Co-Founder. An enthusiast for algorithms coding and design, he has experience in designing and implementing algorithms in the fields of Computer Vision, ML and DL working on big data. Alush obtained a Ph.D. in Computer Vision and Machine Learning algorithms from Bar Ilan University.

Assaf Mushinsky is the startup’s CSO and final Co-Founder. Before Brodmann17, Mushinsky was a researcher and Algorithm Engineer in Samsung Semiconductors and Eyesight, developing algorithms for different Computer Vision tasks. He received a Master’s degree IN Object Detection, Computer Vision and DL from Tel Aviv University.

The $15.2 million raised in funding over four rounds since 2016 has helped Brodmann17’s revolutionary deep-learning vision ADAS solution, designed for automakers, suppliers and aftermarket manufacturers, push its neural-network technology to the fore, creating excitement about the future technology.

*Data taken from Wikipedia

Total
0
Shares
Leave a Reply
Previous Post
Algorithms and Accessibility: Updates from the Advanced Quantum Testbed (AQT)

Algorithms and Accessibility: Updates from the Advanced Quantum Testbed (AQT)

Next Post

Deep Tech Dive #17 | The Past will Present the Future

Related Posts
The Deeptech Insider