Ganesh Balamitran, Product Marketing Manager at Renesas Electronics, demonstrates the company’s latest embedded vision technologies and products at the March 2018 Embedded Vision Alliance Vision Industry and Technology Forum. Specifically, Balamitran demonstrates a Caffe-based CNN for object identification. The demonstration was developed by Renesas partner Uncanny Vision using the Uncanny DLL libraries.
This demonstration is an example of embedded vision implemented at the edge, on a dual-core Cortex A15 @ 1.5GHz (Renesas RZ/G1M MPU). The neural network is trained on a server, but the inference is done at an end-point by the Renesas RZ/G1M processor, which could be implemented in an unconnected, smart, camera device. With devices like the Renesas RZ/G1 MPU, Renesas is bringing embedded vision capabilities to embedded systems.
