Abstract:The process of estimating and counting tree density using only a single aerial or satellite image is a difficult task in the fields of photogrammetry and remote sensing. However, it plays a crucial role in the management of forests. The huge variety of trees in varied topography severely hinders tree counting models to perform well. The purpose of this paper is to propose a framework that is learnt from the source domain with sufficient labeled trees and is adapted to the target domain with only a limited number of labeled trees. Our method, termed as AdaTreeFormer, contains one shared encoder with a hierarchical feature extraction scheme to extract robust features from the source and target domains. It also consists of three subnets: two for extracting self-domain attention maps from source and target domains respectively and one for extracting cross-domain attention maps. 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. We also adopt adversarial learning into the framework to further reduce the gap between source and target domains. Our AdaTreeFormer is evaluated on six designed domain adaptation tasks using three tree counting datasets, \ie Jiangsu, Yosemite, and London. Experimental results show that AdaTreeFormer significantly surpasses the state of the art, \eg in the cross domain from the Yosemite to Jiangsu dataset, it achieves a reduction of 15.9 points in terms of the absolute counting errors and an increase of 10.8\% in the accuracy of the detected trees' locations. The codes and datasets are available at <a class="link-external link-https" href="https://github.com/HAAClassic/AdaTreeFormer" rel="external noopener nofollow">this https URL</a>.

Unsupervised adversarial domain adaptation leaf counting with Bayesian loss density estimation

Attention-based Cross-Layer Domain Alignment for Unsupervised Domain Adaptation

Unsupervised Domain Adaptation For Plant Organ Counting

Background-Aware Domain Adaptation for Plant Counting

Cross Domain Adaptation of Crowd Counting with Model-Agnostic Meta-Learning

AdaTreeFormer: Few Shot Domain Adaptation for Tree Counting from a Single High-Resolution Image

Semantically Guided Bi-level Adaptation for Cross Domain Crowd Counting.

Automatic microscopic cell counting by use of unsupervised adversarial domain adaptation and supervised density regression

Bayesian Uncertainty Matching for Unsupervised Domain Adaptation

Domain Adaptation without Model Transferring

Weakly Supervised Method for Domain Adaptation in Instance Segmentation.

CODA: Counting Objects via Scale-aware Adversarial Density Adaption

Unsupervised Accuracy Estimation of Deep Visual Models using Domain-Adaptive Adversarial Perturbation without Source Samples

Domain-Invariant Adversarial Learning for Unsupervised Domain Adaption.

Model Adaptation: Unsupervised Domain Adaptation Without Source Data.

Multi-source Domain Adaptation with Graph Embedding and Adaptive Label Prediction.

Fine-grained Domain Adaptive Crowd Counting Via Point-derived Segmentation

A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation

A source free domain adaptation model based on adversarial learning for image classification

Self-adaptive Re-weighted Adversarial Domain Adaptation

A Study of Unsupervised Evaluation Metrics for Practical and Automatic Domain Adaptation