Experimental results of different λ with Ls loss function on the LEVIR-CD datasets.

In our proposed SAASNets detection framework, λ = 0.2.

Saved in:

Bibliographic Details
Main Author:	Shuai Pang (4447831) (author)
Other Authors:	Chaochao You (20642598) (author), Min Zhang (111999) (author), Baojie Zhang (511505) (author), Liyou Wang (20642601) (author), Xiaolong Shi (412458) (author), Yu Sun (18463) (author)
Published:	2025
Subjects:	Neuroscience Science Policy Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified Information Systems not elsewhere classified xlink "> interfered shared attention aggregation receptive field limits highly efficient channel head attention module cdd datasets show sharing channel information encode position details detecting different scales building change detection feature aggregation module scale local features irregular object buildings redundant information different stages detection performance feature extractor siamese network siamese features residual strategy position multi global features external factors experimental results encoding stage detailed semantics better accuracy
Tags:	Add Tag No Tags, Be the first to tag this record!

Experimental results of different λ with <i>L</i><sub><i>s</i></sub> loss function on the LEVIR-CD datasets.