Intel Global IoT DevFest IV



Sessions

Using Intel OpenVINO™ To Enhance Quividi's Audience Measurement Solutions


Identification: W7_T3

Quividi, founded in 2006, is a leading company in out-of-home audience measurement solutions based on computer vision. With over 600 customers in 80 countries, Quividi measures over 1 billion viewers per month across diverse retail segments and powers immersive and interactive campaigns for leading brands. All of Quividi's algorithms run on lightweight computing units at the edge for maximum privacy compliance. Over the years, they have evolved from "classic" machine learning techniques to include deep learning solutions powered by Intel® Distribution of OpenVINO™ Toolkit. In this session, Quividi will illustrate the challenges of building a network of measuring units. The presentation will look at some interesting case studies and dive deeper into the OpenVINO™ Toolkit to explore how it can enable a more sophisticated retail experience.

Speaker(s):

Digital Transformation Through the Use of Machine Vision Inside Intel's Assembly and Test Factory


Identification: W7_T1

Digital transformation is a must for several factories to be competitive. This phenomenon is basically to transform traditional factories to one which maximizes the value from data analytics to make a difference. Complex factory opportunities can be viewed from a data analytics point of view and a major part of the solution can be presented from the results of data analytics. Within this scope, there are many different types of data - from text, numerics and images which can be analyzed independently or together. Now, there are different use cases which leverage machine vision, including defect detection and identification. This presentation will contain some of the machine vision use cases within Intel's assembly and test manufacturing.

Speaker(s):
  • Duncan Lee, Manufacturing IT Principal Engineer, Intel Microelectronics Sdn. Bhd.

Revolutionizing Worker Safety with IoT & Innovative Wireless Connectivity


Identification: W8_T2

The Worker Safety solution incorporates smart watches that capture workers’ health data. Leveraging MIOTY connectivity, health and location data of several remote field workers are reliably and securely transmitted to an Advantech edge gateway, where they are relayed to the Hitachi Solutions IoT Service Hub on Microsoft Azure cloud. The solution integrates advanced analytics and data science technology, resulting in a reliable, cost-effective, end-to-end safety offering that is simple to deploy and provides instant visibility to a wealth of actionable operational intelligence.

The solution helps companies meet their Zero Harm and other safety initiatives targets by avoiding risks, decreasing unplanned failures and enabling a safer and healthier work environment. The solution also uses machine learning to continually analyze data to identify the possible incidents and provide a more comprehensive understanding of what occurs in a workplace environment, such as distinguishing normal fluctuations in air pressure vs a potentially problematic anomaly.

Speaker(s):

Face Recognition on the Edge and Cloud


Identification: W8_T1

The face recognition market is expected to reach more than $9 billion by 2024. Deep learning has allowed face recognition to be applied on increasingly more varied, challenging and demanding scenarios, expanding its application from areas like surveillance to new areas such as retail. 

In this presentation, Anyvision’s recognition system - Better Tomorrow - will be introduced. The system, which is being used in tens of thousands of cameras around the world, is capable of processing live video streams in real-time with high accuracy. We’ll explore challenges from different scenarios and examples of our key value. We’ll also discuss how OpenVINO™ has helped us to deploy our solution across a wide range of verticals and use cases with its scalability and unified APIs on a comprehensive selection of silicon devices. We’ll share optimization strategies and experiences learnt on the development of our end-to-end face recognition system to increase performance on different architectures, benchmark results will be shown after that.

Speaker(s):

Breaking the Firmware Barrier Through Open Source Slim Bootloader


Identification: W9_T4

Slim Bootloader (SBL) is an open source platform firmware to initialize platform hardware and boot an operating system through a payload. SBL provides a linear execution flow and relies on Intel® Firmware Support Package provided interfaces to accomplish platform initialization in a simple manner. This simplicity makes platform firmware enabling much easier by democratizing firmware development and encourages more participation from system software development community to cater to various IoT requirements. By taking a modular payload approach, SBL provides scalability to various different requirements by supporting different payloads. These value propositions provided by SBL - including open source with permissive licensing, simple execution flow, scalability and a rich set of features - will translate to lower TTM and TCO for the customers and make SBL an ideal choice for many IoT designs.

Speaker(s):

Quest Global: AI Inferencing on Intel Architecture - How Medical Devices and Surveillance Systems Can Benefit?


Identification: W9_T1b

Quest is an engineering-focused solutions provider to Fortune-500 Technology companies. We build trusted, strategic and global long-term partnership for market leaders across industries. Quest has developed a robust partnership with AIPG and IOTG team. We have developed scalable solutions using OpenVINO™ and Intel® AI in the medical imaging and security and surveillance vertical.

In medical imaging, computer aided detection systems like CT, MRI, X-Ray and Ultrasound help detect tumours and diagnose diseases. These systems can help radiologists and doctors analyze medical images more effectively, resulting in quicker and more accurate diagnosis and efficient workflow. Video analytics has become a very important feature for those employing video surveillance and security systems. AI and machine learning algorithms bring in a new dimension to video analytics solutions making them more reliable and robust.

In this session, we describe the deployment of DL models on IA using two examples – one in the medical domain as a standalone application on edge devices and the other as part of a surveillance platform with DL models residing on an edge device.

Speaker(s):

GeoVision Gets a 24x Deep Learning Algorithm Performance Boost


Identification: W9_T1a

GeoVision is a DSS player in the Top 30 global security companies. Come and see how we work with GeoVision using a set of Intel® Software Tools to optimize computer vision & Deep Learning workloads. Join us to learn the insights that can benefit your own development efforts.

Speaker(s):
  • Joel Lin, Technical Consulting Engineer, Intel Corporation

Transformation of Developer Tools to Address Challenges of Workload Consolidation


Identification: W10_K5

IoT applications are maturing across market segments. Increasing number of systems designers & product managers are trying to figure out how to take advantage of workload consolidation to get a lot more workloads executed with less hardware & complexity. However, a different mindset and tools are required when designing these systems. Traditionally, developers have used tools like editors, debuggers and analyzers which were designed to help build static, monolithic IOT devices. Workload consolidation brings big benefits in performance and operation costs when designed correctly. In this talk, you will get a good understanding of the challenges of designing systems with workload consolidation, and how developer tools have to quickly evolve to help manufacturers and system integrators get to market faster.

Speaker(s):
  • Dinyar Dastoor, Director, Systems and IOT Products, Intel Corporation

The Intel Neural Compute Stick 2: Accelerate Deep Learning Development for Edge Devices


Identification: W11_T1

The Intel® Neural Compute Stick 2 (NCS2) enables rapid prototyping, validation and deployment of Deep Neural Network (DNN) inference applications at the edge. Its low-power VPU architecture enables an entirely new segment of AI applications that aren’t reliant on cloud connectivity. The Intel® NCS2, combined with the Intel® Distribution of OpenVINO™ toolkit, allows developers to profile, tune and deploy Caffe or TensorFlow trained Convolutional Neural Networks (CNN) on low-power applications that require real-time inferencing.

This presentation will provide an overview of the key challenges addressed in delivering AI products and solutions at the edge, example use cases, and the typical journey to develop, prototype and commercialize solutions.

Speaker(s):
  • Lindsay Hiebert, Intel Neural Compute Stick Product Line Owner, Intel Corporation
  • Jason Burris, Intel NCS Developer Outreach Manager, Intel Corporation

The Internet of Things in the Lab: Staging and Testing for the Real World


Identification: W11_T2

Developing and testing software and configuration variants for Industrial Internet of Things (IIoT) applications & systems is a challenge. The systems can be physically large, and often contain 1000s of nodes, which is tough to manage in a physical lab. Testing software that will run across these nodes requires the ability to automate, inspect and control tests, but automating tests across hordes of physical machines is not easy. These challenges can be overcome by using virtual platforms and simulations of wireless networks and the environment. This converts the difficult hardware into software simulations that can be created, configured, and controlled with ease. In this session, we cover the techniques for IoT system simulation and testing that we have developed and discovered over the past year, and how to build scalable testing systems.

Speaker(s):
  • Sean Evoy, Product Line Manager, Wind River
Print Certificate
Review Answers
Print Transcript
Completed on: token-completed_on
Review Answers
Please select the appropriate credit type:
/
test_id: 
credits: 
completed on: 
rendered in: 
* - Indicates answer is required.
token-content
token-speaker-name image
token-speaker-name
token-index
token-content
token-index
token-content
token-index
token-content
token-index
token-content
token-index
token-content
token-index
token-content
/
/
token-index
token-content