Adaptive Label Granularity Selection for Remote Sensing Targets: Balancing Accuracy and Specificity

Remote sensing categories can be organized into a hierarchical label structure, with objects distributed across multiple granularity levels due to substantial differences in resolution and data modalities. However, traditional hierarchical classification performs mandatory leaf-node prediction and cannot stop at intermediate levels when predicting unseen or uncertain targets. In contrast, multi-granularity classification (MGC) can infer targets at any class in the label hierarchy, making it suitable for predicting multi-granularity remote sensing objects. To the best of our knowledge, this is the first study to address the MGC challenge in remote sensing images. Previous studies mainly focus on selecting an optimal threshold for stopping prediction. However, such fixed thresholds cannot flexibly adapt to different tasks and scenarios. This paper proposes an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accuracy-Specificity</i> curve that measures the probability distribution over the label hierarchy and enables the adaptive selection of task-dependent thresholds. The <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accuracy-Specificity</i> curve first ranks hierarchical labels by probability and information content, then evaluates their accuracy and specificity at different thresholds. In top-down prediction along a label tree, increasing label specificity improves information richness but also raises prediction risk, leading to reduced accuracy. Lowering the threshold increases specificity but reduces accuracy, enabling the selection of the appropriate threshold based on the required accuracy or specificity. Moreover, a novel loss function models the semantic distance between the ground-truth node and other nodes in the label hierarchy as a penalty term, adjusting the relative magnitudes of class logits. Experiments on three hierarchical ship datasets show that the proposed method outperforms state-of-the-art methods using conventional and proposed metrics.