Search Results for author:
Found 2 papers, 1 papers with code
AdaTreeFormer: Few Shot Domain Adaptation for Tree Counting from a Single High-Resolution Image
For the latter, an attention-to-adapt mechanism is introduced to distill relevant information from different domains while generating tree density maps; a hierarchical cross-domain feature alignment scheme is proposed that progressively aligns the features from the source and target domains.
TreeFormer: a Semi-Supervised Transformer-based Framework for Tree Counting from a Single High Resolution Image
Automatic tree density estimation and counting using single aerial and satellite images is a challenging task in photogrammetry and remote sensing, yet has an important role in forest management.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
