Instance-Level Few-Shot Learning With Class Hierarchy Mining

Few-shot learning is proposed to tackle the problem of scarce training data in novel classes. However, prior works in instance-level few-shot learning have paid less attention to effectively utilizing the relationship between categories. In this paper, we exploit the hierarchical information to leverage discriminative and relevant features of base classes to effectively classify novel objects. These features are extracted from abundant data of base classes, which could be utilized to reasonably describe classes with scarce data. Specifically, we propose a novel superclass approach that automatically creates a hierarchy considering base and novel classes as fine-grained classes for few-shot instance segmentation (FSIS). Based on the hierarchical information, we design a novel framework called Soft Multiple Superclass (SMS) to extract relevant features or characteristics of classes in the same superclass. A new class assigned to the superclass is easier to classify by leveraging these relevant features. Besides, in order to effectively train the hierarchy-based-detector in FSIS, we apply the label refinement to further describe the associations between fine-grained classes. The extensive experiments demonstrate the effectiveness of our method on FSIS benchmarks. The source code is available here:
https://github.com/nvakhoa/superclass-FSIS