Dear Colleague,![Embedded Vision Summit](https://www.embedded-vision.com/sites/default/files/newsletter/0413/EVSummitLogo_150.jpg)
Highlights of new content on the Alliance website include two
interesting technical articles from Alliance member companies. “A
Quick Guide to Writing OpenCL Kernels for PowerVR Rogue GPUs,” one
in a series of tutorials by Imagination Technologies, gives an overview
of OpenCL programming fundamentals, followed by an explanation of
OpenCL execution on the company’s GPUs. And “Paving
the Way to Self-Driving Cars with Advanced Driver Assistance Systems,”
by Texas Instruments, explores the evolution in vehicle sensing,
intelligence and control that will ultimately lead to self-driving cars.
We also have some great new demo videos. In the
fast-developing world of embedded vision, demos are valuable for
proving the practical feasibility of new approaches, and in sparking
your imagination as to how you might employ these approaches in your
own designs. SarmoTek’s
demonstration of its ADAS software, Texas
Instruments’ demonstration of structured light depth sensing and VectorBlox’
demonstration of various vision algorithms all come from recent
Alliance Member Meetings, while Avnet
Electronics’ demonstration of GigE Vision is the latest in a
burgeoning set of videos from May’s Embedded Vision Summit.
Speaking of the Embedded Vision Summit, be sure to mark your
calendars for next
year’s event, to be held May 2-4, 2016 at the Santa Clara
(California) Convention Center. New advances in deployable computer
vision technology are coming out at rapid rate, and applying these
advances in your new product development can be a foundation for
competitive advantage. Conversely, you might lose ground without them.
Don’t get left behind! Plan to attend Embedded Vision Summit
2016 to learn about the latest practical uses of computer vision
and neural networks. More information on Embedded Vision Summit 2016
will appear on the Alliance website shortly, as well as in future
newsletter editions.
Thanks as always for your support of the Embedded Vision
Alliance, and for your interest in and contributions to embedded vision
technologies, products and applications. I welcome your suggestions
on what the Alliance can do to better service your needs.
Brian Dipert
Editor-In-Chief, Embedded Vision Alliance
|
“How to Create a Great Object Detector,” a Presentation from
MathWorks![MathWorks](https://www.embedded-vision.com/sites/default/files/newsletter/0519/emvideo-youtube-WvTCvHMkZMk_1_125.jpg)
Avinash Nehemiah, Product Marketing
Manager for Computer Vision at MathWorks, presents the “How to Create a
Great Object Detector” tutorial at the May 2014 Embedded Vision Summit.
Detecting objects of interest in images and video is a key part of
practical embedded vision systems. Impressive progress has been made
over the past few years by optimizing object detectors built on
statistical machine learning methods. However, the pre-trained object
detectors available today do not satisfy the increasing diversity of
embedded vision system requirements. This talk will teach you the
basics of creating a robust and accurate object detector. Nehemiah
covers the following topics:
- The importance of good training data sets
- The curse of dimensionality
- Overfitting (why too much training data is not a good
thing), and
- How to select a classifier/detector based on the problem
you are trying to solve.
“Computational Imaging: From Research to Product,” a
Presentation from Personify![Personify](https://www.embedded-vision.com/sites/default/files/newsletter/0519/emvideo-youtube-Xrh0UpS2W-Q_1_125.jpg)
Professor Sanjay Patel shares insights
from his research on computational imaging at the University of
Illinois, and from his experience at Personify, the vision-based
start-up that he co-founded, at the December 2013 Embedded Vision
Alliance Member Meeting.
More Videos
|
Accelerating Machine Learning: Implementing Deep Neural
Networks on FPGAs![Auviz Systems](https://www.embedded-vision.com/sites/default/files/newsletter/0519/Figure1_3_100.jpg)
This article from Auviz Systems discusses
implementing machine learning algorithms on FPGAs, achieving
significant performance improvements at much lower power. Newly
available middleware IP, together with the SDAccel programming
environment, enables software developers to implement convolutional
neural networks (CNNs) in C/C++, leveraging an OpenCL platform model. More
Consumer Willingness to Pay for ADAS Remains Modest as
Interest Rises Sharply![Strategy Analytics](https://www.embedded-vision.com/sites/default/files/newsletter/0519/strategy-analytics_100.jpg)
A recent study from the In-vehicle UX
(IVX) group at Strategy Analytics, surveying consumers in the US,
Western Europe and China, has found some changes in consumer interest
for advanced driver assistance systems (ADAS) which could highlight
potential market shifts in the near future. After years of hesitance,
interest in autonomous safety assistants is finally beginning to show
marked growth. Interest in lane departure warning and park assist is
particularly strong in the US. Park assist now ranks among the top 3
ADAS features consumers would pay more for in the US and Europe. More
More Articles
|