Cross Domain Early Crop Mapping using CropSTGAN
Yiqun Wang, Hui Huang, Radu State
TL;DR
CropSTGAN tackles the problem of cross-domain early crop mapping without target-domain labels by learning a spectral-temporal domain mapper that translates target-domain time-series features toward the source-domain distribution. The framework combines a pre-processor, a CropSTGAN domain mapper (GAN-based) and a TempCNN crop mapper trained on source-domain labels, enabling accurate crop localization in unseen regions or years. Across cross-year and cross-region experiments in the USA and China, CropSTGAN consistently outperforms CropTGAN and STDAN, especially under large distribution shifts, demonstrating strong unsupervised domain adaptation for remote-sensing-based crop mapping. The approach reduces the need for ground-truth data in new regions and can generalize to other crops and land-cover tasks using Sentinel-2 time-series imagery.
Abstract
Driven by abundant satellite imagery, machine learning-based approaches have recently been promoted to generate high-resolution crop cultivation maps to support many agricultural applications. One of the major challenges faced by these approaches is the limited availability of ground truth labels. In the absence of ground truth, existing work usually adopts the "direct transfer strategy" that trains a classifier using historical labels collected from other regions and then applies the trained model to the target region. Unfortunately, the spectral features of crops exhibit inter-region and inter-annual variability due to changes in soil composition, climate conditions, and crop progress, the resultant models perform poorly on new and unseen regions or years. Despite recent efforts, such as the application of the deep adaptation neural network (DANN) model structure in the deep adaptation crop classification network (DACCN), to tackle the above cross-domain challenges, their effectiveness diminishes significantly when there is a large dissimilarity between the source and target regions. This paper introduces the Crop Mapping Spectral-temporal Generative Adversarial Neural Network (CropSTGAN), a novel solution for cross-domain challenges, that doesn't require target domain labels. CropSTGAN learns to transform the target domain's spectral features to those of the source domain, effectively bridging large dissimilarities. Additionally, it employs an identity loss to maintain the intrinsic local structure of the data. Comprehensive experiments across various regions and years demonstrate the benefits and effectiveness of the proposed approach. In experiments, CropSTGAN is benchmarked against various state-of-the-art (SOTA) methods. Notably, CropSTGAN significantly outperforms these methods in scenarios with large data distribution dissimilarities between the target and source domains.