Combining Artificial intelligence, machine learning & deep learning to Embedded systems



With this series now we are moving some another path of Embedded Systems. So what you can think of deep learning, machine learning, and artificial intelligence. I'm sure you have followed some courses, books, articles regarding this subject. Me too followed these courses to learn this subject. We have followed a lot of mathematics modules and stuff with this subject.

So, I would like to share my experience of applying deep learning frameworks to our daily embedded systems applications. That's why I started this tutorial series.

Hardware?

I thought about running a neural network to detect plant behavior using raspberry pi once :-).I tried it with my raspberry pi and totally lagged. Raspberry PI has an ARM processor consist of few cores optimized for sequential serial processing. This will not enough for real-time processing.
A CPU consists of a few cores optimized for sequential serial processing - See more at: http://www.nvidia.com/object/what-is-gpu-computing.html#sthash.7zt3hj7E.dpuf
A CPU consists of a few cores optimized for sequential serial processing - See more at: http://www.nvidia.com/object/what-is-gpu-computing.html#sthash.7zt3hj7E.dpuf
A CPU consists of a few cores optimized for sequential serial processing - See more at: http://www.nvidia.com/object/what-is-gpu-computing.html#sthash.7zt3hj7E.dpuf
A CPU consists of a few cores optimized for sequential serial processing - See more at: http://www.nvidia.com/object/what-is-gpu-computing.html#sthash.7zt3hj7E.dpuf

To do those kinds of applications GPU-accelerated computing is so important.
GPU-accelerated computing is the use of a graphics processing unit (GPU) together with a CPU to accelerate deep learning, analytics, and engineering applications.

The NVIDIA Jetson developer kit gives you everything you need to unlock the power of the GPU for embedded systems applications
 
http://akilacollection.blogspot.com/2016/12/jetson-tx1-embedded-systems-module-for.html 

So this is kind of an introduction. Let's do a some embedded projects applying those theories in the next tutorial.

Comments

Post a Comment

Comment

Popular posts from this blog

IOT Greenhouse project part 1

Image
How to build Simple IOT Greenhouse (Part 1) This this project we are going to control a stable environment to the plant inside the greenhouse.  With using mqtt protocols and latest IOT suff. And also we can monitor plant conditions using digital image processing. So let's start this step by step. Part 1 is all about setting up the sensors and hardware part. So what do you need to this project ? Components 1.) Arduino Uno with wifi shield 2.) Raspberry PI 2 Board and wifi adapter. 3.) DHT 21 (AM2301) Temperature and humidity sensor 4.) Soil Moisture Sensor 5.) Light intensity sensor 6.) Water pump 7.) Raspberry PI Camera or web cam for image processing 8.) 16 in to 4 LCD display Don't be scared i'm going to explain bit by bit how to build this system. First of all you have to design the hardware part. plug the sensor to the arduino in correct way. Simple Arduino Code to get sensor data and display here this code is without the arduino w...
Read more

How to fix FirebaseApp with name [DEFAULT] doesn't exist issue with flutter

Image
How to fix FirebaseApp with name [DEFAULT] doesn't exist issue with flutter This is a very common issue when we try to google authenticate our app with flutter. Please check the following are done in your app. Updated the android build.gradle file Updated the android\app build.gradle file Added dependency to pubspec.yaml Placed the google-services.json file Check Updated the android build.gradle file Add the following line to the android/build.gradle file please add the following line to that file. classpath 'com.google.gms:google-services:4.3.2' // Google Services plugin Check Updated the android\app build.gradle file Add following line to the following location apply plugin : 'com.google.gms.google-services' Check Added dependency to pubspec.yaml firebase_auth : ^0.14.0+5 google_sign_in : ^4.0.7 Placed the google-services.json file make sure you have added google-services.json from the firebase...
Read more

Node.js how to work with Serial ports and display data on a beautiful frontend using Highcharts API Part 2

Image
Now you have some basic idea of the technologies we are using here. So let's get our hands dirty. First, we have to set up the Node js libraries correctly to handle the serial data and open a WebSocket. So simple what we here going to do is getting data from USB and put them into a GUI to visualize our data nicely. So for this tutorial, I'm going to use Arduino as my data source. Piratically this will be your sensor or devices which you are using. Here http://www.highcharts.com/  Highcharts makes it easy for developers to set up interactive charts in their web pages. This has good community support and this is best for our works. 1. First step clone this repository https://github.com/akilawickey/nodejs-serial-frontend/tree/master Here is the Arduino sketch sample.ino which I'm using to emulate data which you can find inside the repository 1: /* 2: * This code will emulate USB data 3: * 4: * 5: */ 6: // the setup function run...
Read more