Machine Learning – solution to complex machine vision tasks (1 day)
Rule-based machine vision has reached its limits in many automation applications. A few of the constraints include changing backgrounds, variable surfaces and objects. Artificial intelligence with machine learning, specifically deep learning, is being used more and more, enabling the reliable implementation of highly complex imaging tasks.
This training will introduce you to the benefits and challenges of conventional machine learning and deep learning, the fundamental differences between the approaches and the possibilities offered by STEMMER IMAGING’s Common Vision Blox (CVB). The training is centered around our tool CVB Polimago. . You will find out about the limits of conventional imaging and experience the fast, intuitive and easy-to-use tool in real-life application examples.
By the end of the course, you will be familiar with the entire machine learning process from the image acquisition to the planning and developing of training sets and the creation, training and evaluation of classifiers. You are invited to bring your application samples and images for an initial evaluation.
People new to the machine vision and artificial intelligence domain. Users with specific tasks and the question if these can be solved with machine learning. Basic machine vision knowledge recommended. No prior machine learning or programming knowledge required.
- Applying conventional rule-based machine vision; the need for machine learning
- When to choose conventional machine vision over machine learning
- Differentiation between deep learning – machine learning
- Theoretical approach: Convolutional Neural Networks (CNN)
- Theoretical approach: CVB Polimago
- Brief introduction to other CVB machine learning tools
Application examples/success stories on CVB machine learning tools Practical hands-on training
- CVB Teach Bench Application
- Image pre-processing and optimisation
- CVB Polimago
- Creating training sets for Search/Classification
- Using created classifiers beyond the CVB Teach Bench application
- Application examples; evaluating classifiers
The training is free of charge.