Adaptive-Labeling for Enhancing Remote Sensing Cloud Understanding
This addresses labeling errors in cloud segmentation for weather and climate science, offering a significant but incremental improvement over existing methods.
The paper tackles the problem of inaccurate labels in remote sensing cloud segmentation by introducing a model-agnostic Cloud Adaptive-Labeling (CAL) approach that iteratively improves training data annotations, resulting in new state-of-the-art performance on multiple benchmarks.
Cloud analysis is a critical component of weather and climate science, impacting various sectors like disaster management. However, achieving fine-grained cloud analysis, such as cloud segmentation, in remote sensing remains challenging due to the inherent difficulties in obtaining accurate labels, leading to significant labeling errors in training data. Existing methods often assume the availability of reliable segmentation annotations, limiting their overall performance. To address this inherent limitation, we introduce an innovative model-agnostic Cloud Adaptive-Labeling (CAL) approach, which operates iteratively to enhance the quality of training data annotations and consequently improve the performance of the learned model. Our methodology commences by training a cloud segmentation model using the original annotations. Subsequently, it introduces a trainable pixel intensity threshold for adaptively labeling the cloud training images on the fly. The newly generated labels are then employed to fine-tune the model. Extensive experiments conducted on multiple standard cloud segmentation benchmarks demonstrate the effectiveness of our approach in significantly boosting the performance of existing segmentation models. Our CAL method establishes new state-of-the-art results when compared to a wide array of existing alternatives.