Category Archive for : ‘Machine Learning’

Introduction

This article shows how to create a PYNQ-DPU application on CorazonAI with the acceleration of the whole AI pipeline flow for video analytics . Specifically this article is a follow-up of the previous article on CorazonAI which can be found here: https://www.makarenalabs.com/pynq-on-corazon-ai-ai-applications-for-all/

AI pipeline and PYNQ

PYNQ is a framework which allows us to use python bindings on the FPGA.

However, having the bindings for python, doesn’t necessarily just mean the possibility of working very easily with our custom IP cores.

In fact, python is a gold standard programming language in computer science and

Introduction

The Corazon-AI board from iWave is very interesting: it has everything you need for an embedded vision application. In the AI world, the time-to-market parameter is very important for a winning application; also currently there is fierce competition, also against the real competitor of the FPGA, the GPU. So we choose to integrate on the Corazon-AI the best ally: the PYNQ framework. PYNQ gives to you the right instruments to create a product-ready application, Python based. You will find a simple tutorial for building the PYNQ framework for the Corazon-AI

Introduction

As we have seen in the last article it’s possible to create and compile a hybrid application on the FPGA system and Kria SOM and use it as a common web application.

At the end of the article there is an incipit about a videogame on FPGA. As weird as that may sound, but we created the first videogame from MakarenaLabs on Kria SOM.

It’s called Krio:“the virtual assistant of MuseBox”.

We have a special board with a high processing power and a little partition dedicated to rendering. It allows you to create

Introduction

In recent days the creation of various web-based application with AI implied the choice of using a cloud architecture This structure can be achieved by creating a REST-API like architecture. The generic structure of this process is the following:

There are sensors (i.e. a camera) which create information;
Secondly cloud receives the information from the collector;
Then cloud performs the AI tasks and return results;
Finally sender uses the results of the inference for its purposes.

However, previous proposed web-based application system has three drawbacks:

First, it needs an high speed internet connection (with its highly

Introduction

In this article we will introduce the reader to the new Xilinx KRIA SOM. So, we will now detail how to take advantage of KRIA SOM in Edge AI application.

Xilinx made a big step towards the Software Engineers with its new Vitis Environment. The new KRIA SOM is another step towards also for ML engineers. KRIA SOM ahs an already built Platform and Image, such that the Engineer just needs to adjust its AI application.

Now, we are going to detail the characteristics of KRIA SOM and what are its features.

KRIA

Introduction

This article introduce a possible architectural design for path planning for Autonomous Guided Vehicles (AGV).
The article proposes a set of functional blocks and their interaction to ensure Real-Time path planning on a Xilinx KRIA System On Module and Xilinx ZCU104.
The article is divided in four main sections:

Modelling of the car-like system
Setup of the system
Trajectory Planning
Proposed Architectural Design

Modelling of the car-like system

The modelization of the car-like is an interesting topic in Analytical Mechanics and Robotics, due the system being underactuated.
A mechanical system is said to be underactuated if its configurations are

Introduction

In this article I am going to explain the concept of predictive maintenance, and why it should be used in any actual application which suffers from time deterioration.

The example application we will focus on is the status report of an electric drive, used in modern applications such as autonomous vehicles or industrial robotics.

The aim of the predictive maintenance is to predict ahead of time eventual faults of the system based on the current status.

We now briefly introduce electrical drives and motor control and then use them to understand how to

FPGA Based Edge AI Gateway with support of AI Frameworks enabling AI Solutions iWave Systems and MakarenaLabs partner to provide comprehensive Edge AI Platforms 

Through a partnership with MakarenaLabs srl, a leader in scientific research in multimedia recognition and feature extractions, iWave Systems is excited to collaborate with MakarenaLabs enabling customers with rugged AI Solutions and edge computing systems. Combining the strong edge computing and FPGA Machine Learning expertise of iWave Systems & MakarenaLabs, through this partnership we aim to deliver an optimized AI platform deploying AI to the edge and

Introduction

For artists, one of the most important problems is to know when their music is reproduced, in order to get royalties from it. 

This article presents a method, which uses Hidden Markov Models (HMM), for understanding by how many songs a specific audio stream is composed of. In Machine Learning we call this particular procedure “segmentation”. The segmentation operation is then defined as a generic operation where a particular object is taken (i.e. an image) and its various components are divided (i.e. foreground and background). In our case the expected output

1 – Introduction

Generally we are used to see articles or demos on deep/machine learning methods just on video applications. In this article we want review and apply the current AI into future trends for robotics, such as learning algorithm.

Let say that there is not a general definition for AI, but the one we selected is that a system is “intelligent” if it achieves a specific goal in a generic situation.

We can define a robotic system as a complex system composed of many heterogeneous components, generally we would have cameras for