![Overview of our model’s architecture. First, one-hot embedding [57,58] is used to embed each topographic factor. Then, we utilize a conditional attention mechanism to learn the different weights of the input factors in relation to archaeological sites and estimate archaeological site locations. The whole model is trained end-to-end.](https://sitereports.site/wp-content/uploads/2023/06/ijgi-12-00238-g005.png)
Overview of our model’s architecture. First, one-hot embedding [57,58] is used to embed each topographic factor. Then, we utilize a conditional attention mechanism to learn the different weights of the input factors in relation to archaeological sites and estimate archaeological site locations. The whole model is trained end-to-end.