The CEVA-CDNN is a comprehensive toolkit that simplifies the development and deployment of deep learning systems for mass-market embedded devices. Tailored and optimized for the CEVA-XM family of imaging and vision DSPs, the CDNN toolkit includes the CEVA network generator, the CDNN software framework, and a CNN hardware accelerator that works together to deliver superior performance while ensuring flexibility with the constantly evolving requirements of machine learning.
Separately, each component of the CDNN toolkit is a powerful enabler of embedded imaging and vision applications. Combined, these pieces deliver an ultimate toolkit to support new network structures and changing layer types of deep neural networks.
CEVA supplies a full development platform for partners and developers based on the CEVA-XM family of DSP cores to enable the development of deep learning applications using the CDNN, targeting any advanced network.
The CDNN toolkit streamlines implementations of deep learning in embedded systems by automatically converting offline pre-trained neural networks to real-time embedded-ready networks for CEVA-XM cores, using the CEVA Network Generator. This enables real-time, high-quality image classification, object recognition, and vision analytics.
- The CEVA Network Generator converts pre-trained neural network models and weights from offline training frameworks (such as Caffe or TensorFlow) to a real-time network model.
- The CDNN2 second-generation neural network software framework accelerates deployment of machine learning in low-power embedded systems.
- The CNN Hardware Accelerator delivers 512 MACs/cycle.
- Coupled with the CEVA-XM family of intelligent vision processors, the CDNN offers significant time-to-market and power advantages for implementing machine learning in embedded systems.
CEVA CDNN2 live demonstration
CDNN2 is CEVA’s 2nd Generation Neural Network Software Framework for Machine Learning that supports Artificial Intelligence Including Google’s TensorFlow
CDNN2 enables localized, deep learning-based video analytics on camera devices in real time. Coupled with the CEVA-XM intelligent vision processor family, CDNN2 offers significant time-to-market and power advantages for implementing machine learning in embedded systems for smartphones, advanced driver assistance systems (ADAS), surveillance equipment, drones, robots and other camera-enabled smart devices.
In this video clip we demonstrate how a pre-trained neural network (“Age Classification”, can be found here: http://www.openu.ac.il/home/hassner/p…) is downloaded from the internet, converted to an embedded-friendly network, and runs on the CEVA-XM4 FPGA platform in less than 10 mins!