Landslide susceptibility mapping by fusing convolutional neural networks and vision transformer

HIGHLIGHTS

SUMMARY

DL can effectively overcome the shortcomings of traditional ML models and efficiently extract deep and intrinsic features of data through multi-layer neural_networks. Convolutional neural_networks (CNN), recurrent neural_networks (RNN), the combination of CNN and RNN, and the combination of CNN and ML have also made good progress in the field of LSM. ViT`s Self-Attention induction bias is weaker than that of CNN, so ViT needs more data to fit the network model. The authors use the Residual Neural_Network (ResNet) model in CNN as the basis and fuse the Multi-Head Self-Attention . . .

 



If you want to have access to all the content you need to log in!

Thanks :)

Username or E-mail:

If you don't have an account, you can create one here.