Comprehensive coverage from basic setup to advanced AI implementations
Learn to optimize AI models with TensorRT for enhanced performance
Develop real-time AI applications using the DeepStream SDK
Description
This course begins with a thorough introduction to Jetson, highlighting its advantages over traditional microcontrollers like the Raspberry Pi. You'll learn how to select the right SD card, flash it effectively, and perform initial configurations to set the stage for advanced developments.
As you progress, delve into installing key AI libraries like OpenCV and PyTorch. Understand their roles in crafting robust AI solutions and enhance their performance with CUDA support. The course offers guidance on fundamental computer vision techniques, enabling seamless image operations, color conversions, and edge detections.
Explore object detection with YOLO and its variants through practical examples. Learn to train custom models like number plate recognition and optimize AI models using NVIDIA's TensorRT for enhanced performance. Dive deep into the DeepStream SDK for real-time video analysis and multi-camera synchronization, vital for security and surveillance applications.
Explore advanced topics like pose estimation, vehicle tracking, and face recognition, all with hands-on projects to reinforce learning. By the end, you will have mastered how to use this powerful platform to push the boundaries of what's possible in AI applications, making you a valuable asset in the tech industry.
What you will learn
Configure and initialize NVIDIA Jetson platforms
Compare Jetson with Raspberry Pi for technological advantages
Install and utilize key libraries like OpenCV and PyTorch on Jetson
Execute basic to advanced computer vision operations using OpenCV
Implement YOLO object detection on custom datasets
Integrate multiple camera inputs using RTSP and ONVIF protocols
Augmented Startups have over 8 years experience in Printed Circuit Board (PCB) design as well in image processing and embedded control. Author Ritesh Kanjee has completed his Masters Degree in Electronic engineering and published two papers on the IEEE Database with one called "Vision-based adaptive Cruise Control using Pattern Matching" and the other called "A Three-Step Vehicle Detection Framework for Range Estimation Using a Single Camera" (on Google Scholar). His work was implemented in LabVIEW. He works as an embedded electronic engineer in defence research and has experience in FPGA design with programming in both VHDL and Verilog. He also has expertise in augmented reality and machine learning in which he shall be introducing new technologies through the medium of video
How can I download a video package for offline viewing?
Login to your account at Packtpub.com.
Click on "My Account" and then click on the "My Videos" tab to access your videos.
Click on the "Download Now" link to start your video download.
How can I extract my video file?
All modern operating systems ship with ZIP file extraction built in. If you'd prefer to use a dedicated compression application, we've tested WinRAR / 7-Zip for Windows, Zipeg / iZip / UnRarX for Mac and 7-Zip / PeaZip for Linux. These applications support all extension files.
How can I get help and support around my video package?
If your video course doesn't give you what you were expecting, either because of functionality problems or because the content isn't up to scratch, please mail customercare@packt.com with details of the problem. In addition, so that we can best provide the support you need, please include the following information for our support team.
Video
Format watched (HTML, MP4, streaming)
Chapter or section that issue relates to (if relevant)
System being played on
Browser used (if relevant)
Details of support
Why can’t I download my video package?
In the even that you are having issues downloading your video package then please follow these instructions:
Disable all your browser plugins and extensions: Some security and download manager extensions can cause issues during the download.
Download the video course using a different browser: We've tested downloads operate correctly in current versions of Chrome, Firefox, Internet Explorer, and Safari.