Welcome to the reComputer Jetson Orin Beginner Guide! Dive deep into the NVIDIA Jetson Orin platform with this comprehensive guide designed to help developers harness Jetson Orin’s powerful AI computing capabilities. By leveraging cutting-edge technology, you will be well-equipped to innovate in AI and robotics. Join us to explore the vast potential of Jetson and set the stage for pioneering developments in the industry!
-
From Beginner to Master:
- Start with the basics and progress to mastering advanced AI applications.
- Modules cover the Jetson Orin software stack, computer vision, video analytics, robotics, and generative AI.
-
Comprehensive Tool Coverage:
- Master NVIDIA's core technologies: CUDA, JetPack SDK, TensorRT, and Deepstream.
- Utilize popular AI frameworks such as PyTorch and TensorFlow.
-
Hands on Industry-Relevant and cutting-edge Projects:
- Build an end-to-end single AI Network Video Recorder (NVR) system in the Computer Vision module.
- Assemble a complete Autonomous Mobile Robot (AMR) in the Robotics module.
- Deploy cutting-edge large language models like Llama 3 and Ollma to create your own chatbot.
-
Step-by-Step Tutorials:
- Receive clear, incremental instructions that guide you from basic programming to the development of complex AI applications on the Jetson platform.
Before beginning, ensure you have:
- Basic knowledge of Linux commands.
- A Jetson device—Seeed reComputer J4012 recommended.
Note: While all Nvidia Jetson Orin-based devices are suitable, ensure your device has at least 8GB of memory.
The reComputer Jetson Orin is a compact yet powerful intelligent edge box that delivers modern AI performance of up to 100 TOPS to the edge. It features an NVIDIA Jetson Orin module, an open-source carrier board, a heatsink, and a power adapter. Key specifications include 4x USB 3.2, HDMI, GbE, M.2 key E for WIFI, M.2 Key M for SSD, RTC, CAN, and a 40-pin connector. Preinstalled with Jetpack, reComputer simplifies development and is ideal for edge AI solution providers focusing on video analytics, object detection, natural language processing, medical imaging, and robotics in smart cities, security, and industrial automation.
Explore a broad range of topics from Jetson platform basics to generative AI deployment:
| Chapter | Content |
|---|---|
| Module 1 | Introduction |
| Module 2 | reComputer Jetson Platform Overview |
| Module 3 | Basic Tools and Getting Started |
| Module 3.1 | Python and Programming Fundamentals |
| Module 3.2 | AI and ML |
| Module 3.3 | Pytorch and TensorFlow |
| Module 3.4 | CUDA |
| Module 4 | Computer Vision Applications |
| Module 4.1 | Overview-of-Computer-Vision |
| Module 4.2 | Real-time-Video-Processing |
| Module 4.3 | Object Detection and Recognition |
| Module 4.3.1 | Train and Deploy YOLOv8 |
| Module 4.3.2 | Deploy YOLOv8 using TensorRT and DeepStream SDK Support |
| Module 5 | Generative AI Applications |
| Module 6 | ROS Robotics |
| Module 6.1 | Introduction to ROS |
| Module 6.1.1 | Overview of ROS and Environment Setup |
| Module 6.1.2 | Quick Experience with HelloWorld for ROS |
| Module 6.1.3 | ROS Architecture |
| Module 6.1.4 | ROS Communication Mechanism |
| Module 6.1.5 | Common ROS Commands |
| Module 6.1.6 | ROS Operation Management |
| Module 6.1.7 | Common Components and Features of ROS |
| Module 6.1.8 | TF Coordinate Transformation in ROS |
| Module 6.2 | ROS Robot Simulation |
| Module 6.3 | Development with Physical ROS Robots |
| Module 6.4 | ROS Project Practice: Advanced Features |
| Module 7 | Algorithm Optimization and Deployment |
| Module 8 | Practical Applications of the Jetson Platform |
| Module 9 | Course Summary and Outlook |
This project is licensed under the MIT License.