The latest release of https://software.intel.com/en-us/openvino-toolkit OpenVINO™ toolkit 2018 R2. Please feel free to share this news with customers.
New Features in the 2018 R2 update include:
- 10 new pre-trained models for retail, pedestrian/vehicle detection, person attributes/emotion recognition, and more (for a total of 20 trained models in the download package)
- Initial ONNX support for 8 models from https://github.com/onnx/models ONNX Model Zoo
- Windows* support for the Intel® Movidius™ Neural Compute Stick (Myriad™ 2)
- OpenCV 3.4.2 features an accelerated deep neural network (DNN) module
- Gold release of the Intel® FPGA Deep Learning Acceleration Suite for real-time AI, enabling CNN workloads.
- Python* API [tech preview], which supports the inference engine
- The Model Optimizer supports the Kaldi* framework - which imports speech models
- Support for Gaussian mixture model and neural network accelerator, enabling offloading neural network inference for speech-related models
Learn more: /discussion/software.intel.com/openvino-toolkit software.intel.com/openvino-toolkit
https://software.seek.intel.com/openvino-toolkit Registration page > Free Download: Linux*, Windows*, Linux for FPGA
- https://techdecoded.intel.io/topics/visual-computing/ Tech.Decoded - online software tools webinars, how-to videos and articles
Background/About OpenVINO™ toolkit: The Open Visual Inference & Neural network Optimization (OpenVINO™) toolkit is a free software toolkit that helps fast-track development of high-performance computer vision and deep learning inference into vision applications. Its part of Intel's end-to-end vision solutions portfolio, and is optimized for multiple Intel® architectures (CPUs, CPUs with integrated graphics, Intel® FPGA, and Intel® Movidius™ VPUs). (This tool was also formerly known as Intel® Computer Vision SDK.)