INTRODUCTION
Artificial Intelligence (AI) and Machine Learning (ML) are nowadays key factors in the world of Advanced Driver Assistance Systems (ADAS), Industrial and Internet of Things (IoT). Due to the systems complexity and safety implications AI & ML require an inseparable combination of processing power and software.
These technologies open entirely new opportunities for the automotive industry to help solve global mobility challenges and NXP is at the forefront in the development of automotive AI & ML solutions with dedicated HW and SW solutions.
The scope of this workshop is to familiarize readers with:
- Deep Learning concepts;
- NXP S32V automotive processors;
- Provide step-by-step prototyping implementation within MATLAB environment
The workshop is a medium to high complexity level and requires only basic understanding of the concepts like neural networks, MATLAB programming and Computer Vision. All the important factors needed for a module completion will be presented as part of this workshop.
For demonstration we are going to use SBC-S32V234 evaluation platform.
Anyhow, in order to avoid any hardware limitations we have designed the workshop’s examples to be generic so that you could use any other design based on S32V234 processor.
METHODOLOGY
Due to subject complexity, we are going to use a mix of Community Articles and Community Videos to explain the concepts behind each topic. Each section will have a dedicated article with all the steps needed to fulfill a functionality and whenever is needed a brief video to explain the concept and/or to show that functionality in real time. All MATLAB functions and accompanying SW will be available for download.
If you are interested in this subject, please bookmark this page or click on Actions/Following In button to get periodic updates. As always, any suggestions for improvement and/or subjects are more than welcome!
WORKSHOP OUTLINE
This workshop is divided in 5 courses that will be released each week giving you plenty of time for interactions and clarifications. The plan of intend is shown below:
C1: HW and SW Environment Setup
- Setting up the hardware: SBC-S32V234 evaluation board, SD-Card for bootup
- Setting up the software: Vision Toolbox, Vision SDK, ARM COMPUTE and Host PC
C2: Introduction into Deep Learning
- What is and what you can do with DL and CNN
- Comparison of various CNN
C3: CNN using MATLAB Simulation Environment
- Using pre-trained networks in simulation
- Use S32V hardware real time data to perform simulation
- Validate algorithm predictions
- Generate C++ code and deploy the CNN on the S32V target
- Execute CNN in real time using S32V processor
- Use MATLAB Deep Learning Toolbox to retrain the CNN
- Validate the retrained CNN on S32V processor
