Abstract:Remote sensing cross-scene classification has always been an important research field, especially in the field of 3-D classification, which is of great significance. Considering the diversity of collection conditions, seasons, and regional styles, deep learning networks well-trained on one source domain dataset tend to suffer from severe performance degradation when applied to other target domain datasets. To tackle the issue, in this article, we propose a new cross-scene classification method, which combines pre-alignment and Shannon entropy constraint to accomplish unsupervised domain adaptive classification (PS-UDA). On the one hand, the pre-alignment employs -paradigm constraint and Laplace matrix to pre-align the features. With the -paradigm constraint, the originally distant features of the source and target domain are constrained to the same sphere surface, and it is easier to make the distribution alignment on the sphere surface. Further, the Laplace matrix is used to map the source and target domain. In this way, similar features of the source and target domain are further aligned, and dissimilar features become discrete from each other. On the other hand, this article employs the Shannon entropy constraint to motivate the network to obtain more high-confidence target domain pseudo-labels. In addition, to fully utilize the unlabeled target domain information, the target domain features are augmented using the adjacency matrix. Experimental results of two cross-scene multispectral point cloud classifications demonstrate that the proposed PS-UDA can effectively mitigate the spectral shift issue in cross-scene multispectral point clouds, achieving state-of-the-art performance.

Unsupervised Domain Adaptation for Crime Risk Prediction Across Cities

DeepOffense: a Recurrent Network Based Approach for Crime Prediction

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

Federated Domain Adaptation via Pseudo-label Refinement

ESCORT: Fine-Grained Urban Crime Risk Inference Leveraging Heterogeneous Open Data

Dynamic road crime risk prediction with urban open data

Unsupervised Domain Adaptation Via Risk-Consistent Estimators

Exploring Transfer Learning for Crime Prediction

Unsupervised Domain Adaptation with Target Reconstruction and Label Confusion in the Common Subspace

Unsupervised Domain Adaptation Via Contrastive Adversarial Domain Mixup: A Case Study on COVID-19

Unsupervised Domain Adaptation: from Simulation Engine to the RealWorld

Unsupervised Domain Adaptation for Cross-Scene Multispectral Point Cloud Classification

Model Adaptation: Unsupervised Domain Adaptation Without Source Data.

An Unsupervised Domain Adaptation Method for Multi-Modal Remote Sensing Image Classification

Unsupervised Domain Adaptation for Object Detection via Cross-Domain Semi-Supervised Learning

Domain Adaptation without Model Transferring

Self-Supervised Boundary Point Prediction Task for Point Cloud Domain Adaptation

A Learnable Self-supervised Task for Unsupervised Domain Adaptation on Point Clouds

Unsupervised Adversarial Domain Adaptation for Cross-Domain Face Presentation Attack Detection

Unsupervised Domain Adaptation for Semantic Segmentation of High-resolution Remote Sensing Imagery Driven by Category-Certainty Attention

HDM-GNN: A Heterogeneous Dynamic Multi-view Graph Neural Network for Crime Prediction