Page
%P
![Loading...](https://link.springer.com/static/c4a417b97a76cc2980e3c25e2271af3129e08bbe/images/pdf-preview/spacer.gif)
-
Chapter and Conference Paper
A Three-Stage Model Fusion Method for Out-of-Distribution Generalization
Training a model from scratch in a data-deficient environment is a challenging task. In this challenge, multiple differentiated backbones are used to train, and a number of tricks are used to assist in model t...