Interest in computer vision at the edge of the network is growing. It helps to solve a range of business problems, including security, access control, and quality control.
The heightened interest in artificial intelligence (AI) for video processing is driven by a number of factors:
- There's been a pivot towards using video data for real-time intervention, instead of after-the-fact analysis.
- Networked IP-based cameras are increasingly available.
- Processing their data at the edge saves bandwidth and accelerates responsiveness.
- Hardware and software portfolios have evolved to make it easier to develop and launch AI applications.
All of these factors have led to an increase in trial deployments. However, the next challenge for companies is to launch their applications at scale. Many times, they don't have the right tools or expertise.
That's why Intel and AWS have created a series of webinars about computer vision at the edge. Across three on-demand webinars, we introduce the tools and technologies for successful computer vision projects.
We hope you'll share the most appropriate webinars with your customers.
Webinar 1: Driving Computer Vision at the Edge
In this webinar, your customers will discover how to accelerate the process from prototype to production, and solution deployment. The webinar outlines the joint Intel and AWS value proposition for computer vision at the edge.
Intel® technologies covered include Intel® Vision Accelerator Design products and Intel® processors, and the Intel® Distribution of OpenVINO™ toolkit. This toolkit enables companies to write their code once and deploy it across a range of hardware devices, taking advantage of available accelerators and processor features.
The webinar showcases the AWS machine learning stack and AWS IoT Greengrass for edge deployment. AWS technologies for video streaming and preparation, machine learning, edge deployment, and reporting accelerate the computer vision project lifecycle. The AWS portfolio includes:
- Amazon Kinesis, Amazon S3, and Amazon Rekognition for video streaming and preparation;
- AWS Deep Learning Amazon Machine Images (AMIs), Amazon SageMaker, and Amazon SageMaker Ground Truth for machine learning;
- AWS Systems Manager, AWS IoT Greengrass, and AWS IoT Device Management for edge deployment; and
- AWS IoT Core, AWS IoT Events, and Amazon Simple Notification Service (SNS) for reporting, notifications, and feedback.
There's a case study showing how CWT Aerospace and Graymatics used computer vision for coastal surveillance. Using an Intel® Vision Accelerator Design product with an Intel® Movidius™ Vision Processing Unit (VPU), an Amazon C4 instance, and an S3 AWS database, the solution can identify swimmers and small boats to help protect the coast from potential threats.
This webinar also outlines how Intel and AWS are making 10% joint investments in customer proofs of concept, with details of the eligibility criteria.
Webinar 2: Edge Devices and Software
In this webinar, we introduce Intel® IoT Market Ready Solutions (Intel® IMRS) and showcase how a successful lab trial was launched into production. Intel IMRS are commercially proven turnkey end-to-end solutions with analytics capabilities that can rapidly scale. The entire solution is available from a single source. Using the Deep Learning Workbench, it is possible to analyze machine learning models to easily tune and optimize performance for Intel® architecture in the cloud.
Viewers learn more about how the enhanced capabilities of AWS IoT Greengrass help with scale deployment and edge device provisioning.
There's a real-world use case showing how defect detection can be carried out using AWS IoT Greengrass. The webinar also discusses how the AWS IoT Events service can be used for interactions that require real-time responsiveness.
Webinar 3: Cloud and AI
AWS EC2 instances based on the Intel® Xeon® Scalable processor family have AI acceleration built in at the hardware level. This enables customers to achieve high throughput for training and inference when used together with optimized frameworks and libraries.1 This webinar includes performance data, showing how AWS instances based on Intel processors perform for machine learning workloads.
The webinar also includes advice on accelerating and streamlining the training process using Amazon SageMaker. Other technologies covered include Ground Truth to speed up labeling, Rekognition to speed up solution building, and the Intel® Distribution of OpenVINO™ toolkit to accelerate neural network models.
Webinar attendees can discover the co-funding opportunities available from Intel and AWS for AI workloads.