alwaysCoding Hacky Hour is our regular workshop designed to share new product releases and exciting content with our developer community.In this Hacky Hour, alwaysAI’s Co-Founder and CTO, Steve Griset, announced the release of edgeIQ 1.7.0 and 1.8.0! He also covered the hardware and software features of NVIDIA Jetson JetPack 4.6.1. with demonstrations of CUDA, TensorRT, and NVDLA. We ended the workshop with a giveaway - a free NVIDIA Jetson Nano for one lucky participant! Watch it on demand and join our future Hacky Hours for more exciting product updates and demos.
In this Hacky Hour we covered:
- edgeIQ 1.7.0
- edgeIQ 1.8.0
- NVIDIA’s new version of CUDA
- NVIDIA’s new version of TensorRT
- NVIDIA’s DLA (Deep Learning Accelerator)
These tools can help you follow along with the demonstration:
GitHub repository: https://github.com/alwaysai/jetpack-46-hacky-hour
Barcode Detection: https://alwaysai.co/docs/edgeiq_api/barcode_detection.html
QR Code Detection: https://alwaysai.co/docs/edgeiq_api/qrcode_detection.html
NVIDIA’s DLA: https://github.com/nvdla/hw
Hacky Hour 34: CUDA & Contours: https://alwaysai.co/blog/cuda-contours
The winner of the free NVIDIA Jetson Nano was chosen based on who could answer this question accurately and the fastest in the chat:
What version of edgeIQ supports Jetson JetPack 4.6?
Answer (Abhijeet): edgeIQ 1.8.0.
QUESTION (Barna): Can I use this for traffic participant detection in a car? That means a city scenario with a lot of pedestrians, cars, and static environments.
ANSWER (Lila Mullany): Yes, you can detect multiple objects as you're describing. TensorRT is another optimization that can increase FPS, which Steve is describing now. If you're interested in discussing it more, you can schedule a demo!
If you haven't already, join our Discord channel to follow up on any questions you have about today's Hacky Hour!
Check out the full video of this Hacky Hour below or click the provided link.