Subscribe
Part of the AvengerCon VII presentations cleared for public release:
Presented by CPT Jack Graham.
The NVIDIA CUDA GPU interface enables desktop-users to build, train, and deploy state-of-the-art machine learning algorithms from their home computer. To demonstrate, we will use the well-known YOLOv5 Object Detection algorithm to train a model that works on niche use-cases. We will go over exactly how the model learns to locate and classify objects, how the underlying network was assembled, and what that means for our data gathering and preprocessing options. With the trained model, we can power the backend of innovative web applications that bring the benefits of machine learning to more users.
Boiler:
AvengerCon is a free security event hosted every fall by Maryland Innovation and Security Institute to benefit the hackers of the U.S. Cyber Command community and the U.S. Army 780th Military Intelligence Brigade. The event is open to all service members and employees of U.S. Cyber Command and Department of Defense personnel supporting cyberspace missions. AvengerCon features presentations, hacker villages, training workshops, and much more.
The event is open to all service members and employees of U.S. Cyber Command and Department of Defense, and related partners supporting cyberspace missions.
The views expressed are those of the presenter, and do not reflect the official position of the 780th Military Intelligence Brigade, U.S. Cyber Command, the Department of the Army, or Department of Defense.
| Date Taken: | 12.01.2022 |
| Date Posted: | 01.04.2023 09:45 |
| Category: | Series |
| Video ID: | 870212 |
| VIRIN: | 221201-O-PX639-207 |
| Filename: | DOD_109398092 |
| Length: | 00:17:11 |
| Location: | US |
| Downloads: | 0 |
| High-Res. Downloads: | 0 |
This work, AvengerCon VII: Creating a Custom Machine Learning Object Detection Model with YOLOv5, by Steven Stover, identified by DVIDS, must comply with the restrictions shown on https://www.dvidshub.net/about/copyright.
