February 14, 2020 - 12:30 PM to 1:30 PM
“Deep Learning for Real-Time Quality Inspection in Manufacturing” - Stephen Welch (VP of Data Science at Mariner)
Driven by rapid advances in deep learning, the third wave of AI has enabled new technologies, new business models, and new ways of solving old problems. Some of the most interesting, economically impactful, and environmentally relevant problems are found in the manufacturing sector. In this talk, we’ll share two case studies from the manufacturing industry detailing the deployment of a real-time GPU-powered Deep Learning vision system for quality inspection. Leveraging a ResNet-based architecture and the Azure IoT stack, we’ve been able to dramatically improve visual inspection performance, delivering a 20-30X performance improvement over existing solutions, allowing our customers to make significant process improvements. We’ll share how our work fits into the broader context of deep learning, address the specific complexities of building, deploying, and maintaining Deep Learning based GPU systems on the edge in manufacturing environments, and detail how these process improvements drive significant business value. Finally, we’ll discuss the open problems in the field, and opportunities to get involved at UNCC.
Stephen Welch is VP of Data Science at Mariner, where he leads a team developing deep-learning based solutions for manufacturing applications. Prior to working with Mariner, Stephen was VP of Machine Learning at Autonomous Fusion, an Atlanta-based autonomous driving startup, where Stephen lead the design, development, and deployment of machine learning algorithms for autonomous driving. Stephen has extensive experience training and deploying machine learning models across a wide variety of domains, including an on-board crash detection algorithm that is now deployed in over 1M vehicles as part of the Verizon Hum product. Stephen strives to not just develop strong technology, but to explain and communicate results in clear and accessible ways – as an adjunct professor at UNCC, Stephen teaches a 60+ person graduate level class in machine learning and computer vision. Finally, Stephen is the author of the educational YouTube channel Welch Labs, which has earned 200k+ subscribers and 10M+ views. Stephen holds 10+ US patents, and engineering degrees from Georgia Tech and UC Berkeley.
Organizer: Min Shin