Flow diagram of material Z classification with transfer learning. The bare material is defined as the source domain, while the coated material is defined as the target domain. (IMAGE)
Caption
As a critical deep learning strategy, transfer learning enables knowledge learned from one domain (the source domain) to be transferred to another related domain (the target domain), effectively mitigating the problem of limited labeled data in the target domain.
Credit
Liang-Wen Chen
Usage Restrictions
Credit must be given to the creator. Only noncommercial uses of the work are permitted.
License
CC BY-NC