On-demand webcast: Selecting the right deep learning tools for your machine vision application

This webcast explores the wide range of deep learning tools currently available for integration in machine vision applications, while advising on how to overcome the challenges associated with their deployment across industries such as semiconductor, batteries, automotive and food packaging. 

Our speakers will identify where exactly deep learning can deliver operational benefit in inspection tasks, and offer guidance on how to select the right deep learning tool based on the vision application under development and materials/objects being inspected.

Attendees will also learn how the latest software can aid them in building and training deep learning models to deliver cost-effective vision inspection applications, demonstrating how accessible it is for anyone to create such models in order to achieve optimal performance.
 

Who should download?

• Automation engineers
• System integrators
• Engineering leaders
• Machine-vision users
• Operations leaders
• Quality-control personnel

Speakers

Christian Eckstein, Business Developer and Partner Manager at MVTec Software
Deep learning in industrial applications – tools and best-practices

Deep learning is a technology widely used across various industries and applications. However, industrial deployment comes with a unique set of challenges. From speed to reliability, hardware requirements and the lack of high-quality datasets, industrial requirements are wide ranging and call for the right tools and processes. 

In this short presentation Eckstein will provide an overview of the basic deep learning methods used in industrial machine vision, the unique challenges associated with them and demonstrate how a typical deep learning pipeline aims to address them.

--

Antoine Lejeune, Vision Software Engineer, Euresys

Euresys Deep Learning tools: A data centric approach for managing your deep learning projects from annotation to deployment

Lejeune will explore the applications in which deep learning can bring operational benefits to inspection tasks, and how Euresys tools can help you create, track and optimise deep learning models.

He will review classic deep learning tools for classification, segmentation and localisation and compare them to a traditional image processing approach – focussing on the pros and cons of each method and how to select the right deep learning tool based on the nature of your data.

Lejeune will also highlight how data is at the heart of deep learning applications and how Euresys tools integrate them into the workflow. He’ll cover image annotations and the evaluation of trained model performance, showing how to iteratively improve the annotations and performance of your models using Euresys tools. Finally, he will also discuss how to maximise your model's inference speed for deployment and production.

--

Hongsuk Lee, CEO, Neurocle
Maximising model performance in one click: Auto deep learning vision inspection software

In his current role as CEO of Neurocle, Lee is focused on the development of the firm’s flagship product, an auto deep learning vision software designed for universal accessibility. He will be joining us to discuss the software and its versatile applications across the semiconductor, batteries, automotive and food packaging industries.  

Lee will explore how anyone, using the right software, can efficiently undertake a cost-effective vision inspection project by streamlining the training and evaluation of deep learning models. For example, Neurocle’s Neuro-T auto deep learning vision software reduces labelling and training time, giving users the flexibility to select from six deep learning models that align with their specific project needs. Lee will provide a brief overview of Neuro-T’s application across the above industries, presenting an excellent opportunity for domain experts to discover how accessible it is for anyone to create deep learning models and achieve optimal performance.