With AI models based on a convolutional neural network (CNN) enabling valuable insights from video or image data, and with the availability of low-power neural network accelerator hardware - the exciting opportunities for using vision at the edge are now thriving!

• Applications that use video analytics at the edge
• Deep learning models for inferencing
• Composition of a typical video pipeline
• Preprocessing, inference and post-processing requirements for video analytics
• Realizing video analytics using Python, C++ API’s or using a Linux Gstreamer based pipeline
• Optimizing models for embedded hardware with neural network accelerators
• Deploying the video analytics solution

The presentation will include a working example of how video analytics can be deployed using an Aquila AM69 board from Toradex, featuring an AM69 processor from Texas Instruments.

Note: The video recording of the webinar will be made available here shortly.