Develop Your Edge AI Skills with Free Online Tools

Table of Contents:

1. Introduction
2. The Importance of AI in Edge Devices
3. Getting Started with AI Hands-On Code Development
4. Overview of Jacinto AI Technology
5. The Hello World Program in the AI World
6. The Complexity of AI Development
7. Simplifying AI Development with TI's Jacinto Architecture
8. Deep Learning and Neural Networks
9. Selecting and Optimizing Models for AI Applications
10. Using TensorFlow Lite for AI Development
11. Deploying AI Models onto Edge Devices
12. Real-World Use Cases for AI in Edge Devices
13. Advanced Development and Optimization Techniques
14. Resources and Further Reading
15. Conclusion

The Importance of AI in Edge Devices

Artificial Intelligence (AI) has become increasingly important in edge devices, such as automotive and industrial applications. These devices are now equipped with Jacinto AI technology, which brings AI capabilities to a wider range of markets. In this webinar, we will explore the fundamentals of AI hands-on code development, with a focus on embedded AI development. We will also introduce the Jacinto architecture and demonstrate how TI is making it easier for developers to use popular machine learning frameworks for embedded AI development.

Getting Started with AI Hands-On Code Development

If you are new to AI hands-on code development, this webinar is the perfect opportunity to learn the fundamentals. In the next 45 minutes, we will provide you with all the necessary tools and code examples to develop your first Hello World program. Whether you are a beginner or an experienced developer, this webinar will help you get started with AI development and provide insights into different software frameworks and tools commonly used by developers.

Overview of Jacinto AI Technology

In this section, we will provide an overview of Jacinto AI technology and its applications in automotive and industrial settings. Our speakers, Srik Gurrapu, Manisha Agrawal, Katrina Tuazon, and Mukul Bhatnagar, will introduce themselves and explain the goals of this webinar. They will also be available throughout the presentation to answer any questions you may have.

The Hello World Program in the AI World

The Hello World program is a tradition in the world of programming, and we are taking it to the next level in the AI world. Instead of a simple "Hello World" message, we will demonstrate how to do some fancy things, such as image pre-processing and object recognition. We will show you how to take an image, read and pre-process it, automatically determine the objects in the image, and output the results in an easy-to-understand format. This program will serve as a starting point for your AI development journey.

The Complexity of AI Development

AI development can be complex, but at TI, we are committed to making it simpler. We understand that the Hello World program in the AI world is much more complex than traditional programming. However, we have developed tools and frameworks that simplify the process. In this webinar, we will guide you through all the necessary steps to develop AI applications, from running the code on a PC to optimizing it for embedded edge devices.

Simplifying AI Development with TI's Jacinto Architecture

TI's Jacinto architecture is designed to simplify AI development and enable AI in lower power and low complexity devices. The architecture includes a deep learning accelerator, imaging and vision accelerators, and an ARM A72 core. These components work together to provide a seamless AI development experience, making it easy to use popular machine learning frameworks and get your embedded AI development started.

Deep Learning and Neural Networks

Deep learning is at the core of AI development. It involves creating models that are trained on a known dataset and can automatically recognize objects in new data. In this section, we will discuss the basics of deep learning and how it applies to AI development. We will also introduce popular datasets and explain how to use them in your AI applications.

Selecting and Optimizing Models for AI Applications

Selecting the right model for your AI application is crucial for achieving accurate results. In this section, we will discuss different models and their trade-offs in terms of accuracy and operations. We will also introduce TI's model zoo, which includes verified models that can be used in production. Additionally, we will provide a tool to help you choose the best model based on your accuracy and speed requirements.

Using TensorFlow Lite for AI Development

TensorFlow Lite is a lightweight version of TensorFlow, specifically designed for embedded edge devices. It offers support for industry-standard frameworks and can be easily integrated into your AI development workflow. In this section, we will explain how to install TensorFlow Lite and demonstrate its usage in developing AI applications. We will also discuss other useful software libraries, such as OpenCV and NumPy, that can enhance your AI development experience.

Deploying AI Models onto Edge Devices

Once you have developed and optimized your AI model, the next step is deploying it onto edge devices. In this section, we will explain the three steps involved in adding AI to your system: running the code on a PC, porting it to an ARM processor inside the Jacinto SoC, and enabling hardware acceleration. We will also demonstrate how to deploy the optimized model onto your target processor and run it on top of Linux.

Real-World Use Cases for AI in Edge Devices

AI has vast applications across various industries, from robotics to factory automation. In this section, we will explore different use cases where AI can be implemented on edge devices. For example, AI-enabled cameras can be used for automatic object detection and person recognition, while AI-enabled robots can perform tasks such as hazard detection and collision avoidance. We will discuss the unlimited possibilities of AI in edge devices and highlight the benefits of AI-enabled systems.

Advanced Development and Optimization Techniques

In this section, we will dive deeper into advanced development and optimization techniques for AI applications. We will discuss topics such as custom model development, transfer learning, and optimization of model weights. We will also provide resources and further reading for developers who want to explore these topics in more detail. Additionally, we will introduce TI's comprehensive software SDK and demonstration tools that can help you accelerate your AI development journey.

Resources and Further Reading

To help you on your AI development journey, we have compiled a list of resources and further reading materials. These include links to the model zoo, SDK documentation, and additional tools and libraries for AI development. We encourage you to explore these resources and continue learning about AI in edge devices. The possibilities are endless, and we are here to support you every step of the way.

Conclusion

In this webinar, we have covered the fundamentals of AI hands-on code development and demonstrated the power of AI in edge devices. We have explored TI's Jacinto AI technology and its applications in automotive and industrial settings. We have also provided a step-by-step guide to developing your first Hello World program in the AI world. With TI's comprehensive software offering and powerful hardware accelerators, you can take your AI development to the next level. We encourage you to continue exploring AI in edge devices and leverage the resources and tools available to you. Thank you for joining this webinar, and we look forward to helping you succeed in your AI development journey.