Master Class: Accelerate Deep Learning Inference using Intel OpenVINO™ Toolkit
Why attend?
These workshops or master classes are well selected for IoTForum's goal of Raising the IoT Quotient of India. In these days of rapid change, startups need to prepare for the future rather than be caught with outdated technologies and old school business models. Some links to Thought Leadership are illustrated in the main page of the IoTForum Workshops. We have kept descriptions of the context and courses brief to allow rapid scanning but do click through the hyperlinked portions, explore topics and blogs at IoTForIndia to get more depth and send us a query or join our Telegram Group and engage. Checkout this video to get a bird's-eye view of the workshops.
Theme
Edge computing has been gaining popularity as latency of cloud-based command and control as well as #calldrop raise issues. AI or Machine Learning inference on edge has a strong business case. It works faster, takes less power and is more reliable as network issues are bypassed. To make this possible a new generation of AI chips are under development. Current approach takes FPGA blocks and special compilers that reduce common neural net models to a small size and run it with much less power and efficiently. GPU are not needed at the edge.
Intel has merged many capabilities into its new toolkit. The OpenVINO toolkit offers software developers a single toolkit for applications wanting human-like vision capabilities. It does this by supporting deep learning, computer vision, and hardware acceleration with heterogeneous support, all in a single toolkit
Organization
Intel is one the largest semiconductor vendor with CPU and other chips. It acquired Nervana and eASIC and offers Movidus USB based neural compute stick.
OpenVINO offers specific capabilities for CNN-based deep learning inference on the edge. It also offers a common API that supports heterogeneous execution across computer vision accelerators: CPUs, GPUs, Intel® Movidius™ Neural Compute Stick, and FPGAs on Linux and Windows. Supports all object detection models on TensorFlow Model Zoo, all public models on Open Neural Network Exchange (ONNX)
Speaker
|
Hemanth Kumar G is an Embedded IOT Technical Consulting Engineer from Developer Products Division, Intel, India. Presently focusing on Intel System Studio tools like Intel System Debugger, VTune Amplifier, ICC; OpenVINO Toolkit, Intel Optimized Tensorflow, Intel Caffe & Media SDK from Intel. He possess more than 11 years of vast experience in academia and industry. His expertise include IOT, AI&ML, Computer Vision, Computer Graphics & Multimedia, Heterogeneous Parallel Computing. |
| Krishnakumar Shetti (Krishna) is a Technical Consulting Engineer with the Intel® Core and Visual Computing Group (CVCG), specializing on Intel® products and tools related to Embedded/IoT space. His experience includes system debug, power/performance analysis and edge analytics on various Intel® platforms. He has a Master’s Degree in Embedded Systems from Manipal University. |
Takeaways
- Using the toolkit to deploy a neural network and optimize models
- Intel Deep learning Inference H/w Platforms
- The Intel® Deep Learning Deployment Toolkit—part of OpenVINO™—including its Model Optimizer (helps quantize pre-trained models) and its Inference Engine (runs seamlessly across CPU, GPU, FPGA, and VPU without requiring the entire framework to be loaded)
- How the Inference Engine lets you utilize new layers in C/C++ for CPU and OpenCL™ for GPU
Audience
- Developers with good understanding of Neural Networks and appreciation of edge devices and microcontrollers.
- Architects, Product Managers for embedded solution
Schedule
| 10.00 AM - 01.00 PM | Presentation covering openVINO |
Venue
Diamond District, Lower Ground Floor, DD3, HAL Old Airport Rd, ISRO Colony, Domlur, Bengaluru, Karnataka 560008Limited Paid parking available
Queries
