Artificial Intelligence(AI) at Edge

In this particular article , I am discusing AI based Ecosystem stm32cube.AI . This includes the quick tour of installation procedure.

In my last article of Embedded Career , I tried to explain the role of Embedded in AI in a layman approach. Here  we can have further insight into that.

Why AI is required ?

Recently I came across one query in a forum ” Why we need AI ? What AI can do which human can’t ? ”  At a first glance, this seems a naive question. But being a computer professional , we should not forget the value of 0.

When I learned driving  using manual gears , I used to ask my trainer about different conditions when a particular gear was required. I was trying to adopt it in a digital fashion . Later I realized that it is more than just following straight rules.  After going through training  and learning from mistakes,  One can manage driving car.  However one can’t guarantee a safe journey due to various factors related to person behind the wheel and this leads to road accidents.

AI has capability to make a  vehicle  driver-less or self-driving or autonomous or robot vehicle.  Although human can drive a car but need a specific skill and experience.

What is AI ?

Lets try to understand AI with driver-less car.  As you know, the driver action depends mainly on sound , vision and brain. Now one can think of various activities related to these such as speech recognition, object detection , Image capture , Image processing etc  at first level where MCU can be used.

So if you are getting started in AI with embedded , You may work on one of the sound or vision related activity.

What is the relationship among AI, ML and DL  ? 

Machine Learning(ML) means   Machine learns itself . It is similar to learning by human based on experiences. In case of machine ,  this will be based on statistics and will be used for prediction.

Deep Learning(DL) passes a bulk of data through layered structure consists of different algorithms to simulate human brain. This can be used for Voice recognition , Image recognition etc .

AI Ecosystem from Semiconductor  Companies

NxP has AI and ML portfolio for Automotive, Industrial and IoT application.  I am going to look into ST related ecosystem as our focus is on STM32 cube family.

You may want to go through a quick tour of overall setup with STM32cube,AI .

I have posted one at  at Artificial Intelligence@Edge

If you are planning to use this , Please visit STM32cubeAI Getting Started

Thanks for reading till end. We will keep posting  Embedded specific important topics.  If you find useful tips at Embedkari, Please subscribe  Embedkari so that you don’t miss any interesting topic .





Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.