See All Categories

Learn More



Current Page

  1. Home
  2. Forum
  3. Press

Press

Press

Forum Details
Title [IoT Korea 2016] Neurocomes presents a small module based on deep-learning, the Deep Runner
Posted by neurocoms (ip:)
  • Date 2016-11-22
  • Recommend Recommend
  • View 124
  • Rate 0points




Neurocomes offers a deep-learning based, low power-consuming small module that can serve as a visual sensor for IoT network by detecting various objects in real-time based on its unique single chip and low power-consuming hardware technology which will replace GPUs in the deep learning hardware market.

The Deep Runner by Neurocomes is a universal IoT visual sensor module using deep learning technology to detect learned objects from video footage. It is difficult to use GPUs for an IoT environment since they consume hundreds of watt for deep learning processing. However, with the ‘Neuron Machine,’ Neurcomes’ unique neural network hardware technology, this module can process the GoogleNet, a deep learning algorithm, that boasts a high detection rate with energy consumption of just few watts, and can be applied for various IoT products.


Click the link below to read the full article.

http://www.hellot.net/new_hellot/magazine/magazine_read.html?code=202&sub=004&idx=30847





Attachment
Password * Please enter password to remove this post.
Admin Spam Not Spam
List Delete Edit Reply
Edit Comment

Password

Edit Cancel

/ byte

Add Comment

Add a commentNamePasswordView Reply

Submit

/ byte

* Please type without any spaces.(case sensitive)

Comments can be added by Member only.