English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Transfer learning improves the performance of deep learning models byinitializing them with parameters pre-trained on larger datasets. Intuitively,transfer learning is more effective when pre-training is on the in-domaindatasets. A recent study by NASA has demonstrated that the microstructuresegmentation with encoder-decoder algorithms benefits more from CNN encoderspre-trained on microscopy images than from those pre-trained on natural images.However, CNN models only capture the local spatial relations in images. Inrecent years, attention networks such as Transformers are increasingly used inimage analysis to capture the long-range relations between pixels. In thisstudy, we compare the segmentation performance of Transformer and CNN modelspre-trained on microscopy images with those pre-trained on natural images. Ourresult partially confirms the NASA study that the segmentation performance ofout-of-distribution images (taken under different imaging and sampleconditions) is significantly improved when pre-training on microscopy images.However, the performance gain for one-shot and few-shot learning is more modestwith Transformers. We also find that for image segmentation, the combination ofpre-trained Transformers and CNN encoders are consistently better thanpre-trained CNN encoders alone. Our dataset (of about 50,000 images) combinesthe public portion of the NASA dataset with additional images we collected.Even with much less training data, our pre-trained models have significantlybetter performance for image segmentation. This result suggests thatTransformers and CNN complement each other and when pre-trained on microscopyimages, they are more beneficial to the downstream tasks.

Transfer Learning for Microstructure Segmentation with CS-UNet: A Hybrid Algorithm with Transformer and CNN Encoders