Abstract:Appearances and qualities of remote sensing images are affected by different atmospheric conditions, quality of sensors, and radiometric calibrations. This heavily challenges the generalization ability of a deep learning or other machine learning model: the performance of a model pretrained on a source remote sensing data set can significantly decrease when applied to a different target data set. The popular generative adversarial networks (GANs) can realize style or appearance transfer between a source and target data sets, which may boost the performance of a deep learning model through generating new target images similar to source samples. In this study, we comprehensively evaluate the performance of GAN-based image-level transfer methods on convolutional neural network (CNN) based image processing models that are trained on one dataset and tested on another one. Firstly, we designed the framework for the evaluation process. The framework consists of two main parts, the GAN-based image-level domain adaptation, which transfers a target image to a new image with similar probability distribution of source image space, and the CNN-based image processing tasks, which are used to test the effects of GAN-based domain adaptation. Second, the domain adaptation is implemented with two mainstream GAN methods for style transfer, the CycleGAN and the AgGAN. The image processing contains two major tasks, segmentation and object detection. The former and the latter are designed based on the widely applied U-Net and Faster R-CNN, respectively. Finally, three experiments, associated with three datasets, are designed to cover different application cases, a change detection case where temporal data is collected from the same scene, a two-city case where images are collected from different regions and a two-sensor case where images are obtained from aerial and satellite platforms respectively. Results revealed that, the GAN-based image transfer can significantly boost the performance of the segmentation model in the change detection case, however, it did not surpass conventional methods; in the other two cases, the GAN-based methods obtained worse results. In object detection, almost all the methods failed to boost the performance of the Faster R-CNN and the GAN-based methods performed the worst.

Cross-Domain Hyperspectral Image Classification Based on Generative Adversarial Networks

Cross-Scene Hyperspectral Image Classification Based on Cycle-Consistent Adversarial Networks.

X-Gacmn: An X-Shaped Generative Adversarial Cross-Modal Network With Hypersphere Embedding

Cross-Domain Hyperspectral Image Classification Based on Graph Convolutional Networks

Building Cross-Domain Mapping Chains from Multi-CycleGAN for Hyperspectral Image Classification

Cross-domain attention network for hyperspectral image classification

Two-Stage Domain Alignment Single-Source Domain Generalization Network for Cross-Scene Hyperspectral Images Classification

Domain‐invariant Attention Network for Transfer Learning Between Cross‐scene Hyperspectral Images

A GAN based method for cross-scene classification of hyperspectral scenes captured by different sensors

Dictionary Learning-Based Feature-Level Domain Adaptation for Cross-Scene Hyperspectral Image Classification

Cross-Domain Sentiment Classification with Attention-Assisted GAN

Hyperspectral Images Single-Source Domain Generalization Based on Nonlinear Sample Generation

Cross-Domain Hyperspectral Image Classification Based on Transformer

Learning a Deep Structural Subspace Across Hyperspectral Scenes with Cross-Domain VAE

Cross-domain Residual Deep NMF for Transfer Learning Between Different Hyperspectral Image Scenes.

Class-Aligned and Class-Balancing Generative Domain Adaptation for Hyperspectral Image Classification

Generative Adversarial Networks and Probabilistic Graph Models for Hyperspectral Image Classification

Spectral–Spatial Adversarial Multidomain Synthesis Network for Cross-Scene Hyperspectral Image Classification

Cross-Scene Feature Selection For Hyperspectral Images Based On Cross-Domain Information Gain

Effective Visual Domain Adaptation via Generative Adversarial Distribution Matching

Evaluating generative adversarial networks based image-level domain transfer for multi-source remote sensing image segmentation and object detection