See All Categories

Learn More



현재 위치

  1. Home
  2. Product

About Deep Runner

The Deep Runner is a low-power universal device with the latest deep learning algorithms recognizing any object with a high accuracy rate. It is small, low-energy consuming but smart, just like a human brain.

 Model characteristics

 What are the major features of the Deep Runner?

The table below shows supported deep learning algorithms and their speed. Users can choose an appropriate algorithm per requirement.


Speed of each deep learning algorithm (NC601)

 The function of the Deep Runner

Classification and detection of the Deep Runner

The Deep Runner not only classifies an object but also detects the location of the object in videos.

Multi-domain Detection Feature

The Deep Runner loads multiple models (trained knowledge) on the memory and instantly switches back and forth to detect different objects.
Changing domains does not take time at all.

One-stop Deep Learning Training Software

A deep learning software program and complex programming are not required.
Train your deep learning system just by providing images.

Automatic image collection for training

The Deep Runner collects video scenes used for detecting objects. The key to high accuracy lies in securing quality training data.
The Deep Runner helps you to build a data set for training.

 Single configuration,Integrated configuration,Client/server structure

Single configuration

The Deep Runner can be used independently. It notifies the user when it detects a target object. It is easy to install. It can help soldiers at a guard post by the front line or detect a natural disaster or abnormal situation in advance.

Integrated configuration

Connect the Deep Runner to a service system via Ethernet port and send the detection result in real-time using simple packets. The service system can be a host system that is already operating, or a small system such as Raspberry Pi and Arduino. The Deep Runner can serve as a visual sensor for a large network when it is connected to an IoT platform through the service system.

Client/server structure

The connection structure is the same, but the service system keeps the source of videos and requests the Deep Runner for detection whenever it is needed and get the result immediately.

 Applications of the Deep Runner

- Security system -

- Apartment security camera surveillance -

- Monitoring legal violation -

- Early detection of disasters and pests -

- Quality inspection in a factory -

 Who can use it?

 System developer

 Entrepreneurs implementing new ideas

 DIY (do it yourself) developers (such as a robot)

 Training kit for a deep learning study

 Comparison of costs for establishing deep learning systems

 When can I buy it ?

The Deep Runner is going to be launched in July 2017.