DEV Community

Discussion on: Dec. 6, 2019: What did you learn this week?

Collapse
 
jinglescode profile image
Jingles (Hong Jing)

I learn about this cool machine learning architecture call UNet++, that aims to improve segmentation accuracy, with a series of nested, dense skip pathways.

Experiment it with PyTorch on my medical images dataset, it is small (usually medical dataset are small) yet it is able to perform.

Wrote about it in this article to help me digest the paper for future reference and for sharing with all of you!