Cadence Demonstration of Deep Learning-based Image Classification Using XNNC-generated Code скачать в хорошем качестве

Cadence Demonstration of Deep Learning-based Image Classification Using XNNC-generated Code

For more information about embedded vision, including hundreds of additional videos, please visit https://www.embedded-vision.com. Pulin Desai, Product Marketing Director at Cadence, demonstrates the company's latest embedded vision technologies and products at the January 2018 Consumer Electronics Show. Specifically, Desai demonstrates the company’s Tensilica® Vision P6 DSP in an on-device AI (artificial intelligence) application. A 22nm FDSOI process-based SoC that integrates the Vision P6 is processing code for image classification, generated by the XNCC (Xtensa Neural Network Compiler). The XNNC can generate code targeting the Cadence AI processor core from neural networks trained using TensorFlow and Caffe. The platform used for the demonstration can also be used by Cadence's customers and partners for rapid software development.