Synthetic Training Images for Semantic Segmentation of Dendrites via Machine Learning

Jiwon, Yeom; Stan, Tiberiu; Seungbum, Hong; Peter, Voorhees

simulation machine learning metals and alloys convolutional neural network semantic segmentation solidification serial sectioning Phase-field model microscopy microstructures

Organizations

MDF Open

Year

2020

Source Name

jiwonyeom_pfimages_learning

License

CC-BY 4.0

Contacts

Jiwon Yeom (jiwon58@kaist.ac.kr) KAIST Tiberiu Stan (tiberiu.stan@northwestern.edu) Northwestern University

DOI

10.18126/fj1u-cvn9 View on Datacite

Get the Data

Synthetically generated phase-field images for training and validation sets used to train dendrite neural nets. Test image is coming from X-ray tomography for measuring the performance of neural nets.