AI Driven Visual Inspection Systems

Overview

Automated visual inspection is crucial for efficient quality control in production systems and is widely used in manufacturing and infrastructure industries. Recent advancements in deep learning have introduced new tools for automated visual inspection tasks with high accuracy and robustness. These new methods enable the detection of defects without relying on failure data during training, offering unprecedented capabilities for detecting flaws on different manufactured surfaces.

In this session, we will introduce state-of-the-art approaches for visual inspection especially anomaly detection and object detection methods that are included in Computer Vision Toolbox Automated Visual Inspection Library. Through a couple of user case studies, you will learn AI DevOps workflow, some tips and techniques that can be used in your project. You will see how to integrate a visual inspection algorithm into a larger system.

Highlights

• Preprocessing and augmenting image data
• Anomaly detection and object detection
• Overview of DevOps AI Workflow with Apps for low-code AI
• Automatically generate code and deploy to embedded targets.
• Validating and operationalizing models

About the Presenter

Jayanth Balaji Avanashilingam works as a Senior Application Engineer at MathWorks in Artificial Intelligence.
He primarily focuses on areas of Data Analytics for the application involving with Time-Series data.
Jayanth has around 8 years of research and industrial experience, where he was working developing AI/ML/DL solutions for various application areas, such as time-series data modeling, optimization, computer vision and Natural Language Processing. Prior to joining MathWorks Jayanth was working as Senior AI Engineer at Impact Analytics, Bangalore.

Ramanuja Jagannathan works as Senior Application Engineer at MathWorks . He works on modelling, simulation, control design and simulation deployment for mechatronics applications. He has experience working in customer projects involving design of mechatronic machine design, robotic manipulator, reinforcement learning based controller and digital twin deployment. Before joining MathWorks, he was as a Senior Engineer at Larsen and Toubro working in project planning and control of Railway projects. His academic background is in Process Control, and he did master’s programme at National Institute of Technology, Tiruchirappalli, India.