Our team members Martin Mรผller, Marie Stiefel and Bjรถrn Bachmann as well as our managing directors Dominik Britz and Frank Mรผcklich have written a review paper on the current state of the art of AI-based microstructure analysis at the invitation of the Metals Editorial Office. โ
Their article “๐๐ฏ๐๐ซ๐ฏ๐ข๐๐ฐ: ๐๐๐๐ก๐ข๐ง๐ ๐๐๐๐ซ๐ง๐ข๐ง๐ ๐๐จ๐ซ ๐๐๐ ๐ฆ๐๐ง๐ญ๐๐ญ๐ข๐จ๐ง ๐๐ง๐ ๐๐ฅ๐๐ฌ๐ฌ๐ข๐๐ข๐๐๐ญ๐ข๐จ๐ง ๐จ๐ ๐๐จ๐ฆ๐ฉ๐ฅ๐๐ฑ ๐๐ญ๐๐๐ฅ ๐๐ข๐๐ซ๐จ๐ฌ๐ญ๐ซ๐ฎ๐๐ญ๐ฎ๐ซ๐๐ฌ” (Metals 2024, 14, 553) deals with detailed use cases using the example of earlier publications. ๐
The paper pays particular attention to the quantity, quality and variance of the training data as well as the question of where the ground truth required for ML comes from, which is usually not sufficiently discussed in the literature. ๐
If you are interested in AI-based microstructure analysis and want to get up to date, you should definitely not miss this review paper! โจ