Fine-Grain Few-Shot Vision via Domain Knowledge as Hyperspherical Priors
This addresses the challenge of distinguishing very similar classes in computer vision for applications like fine-grained image recognition, representing an incremental improvement over existing methods.
The paper tackled the problem of few-shot fine-grain classification by incorporating domain knowledge as hyperspherical priors in prototypical networks, achieving top results on benchmark datasets with 5x speedups in training time.
Prototypical networks have been shown to perform well at few-shot learning tasks in computer vision. Yet these networks struggle when classes are very similar to each other (fine-grain classification) and currently have no way of taking into account prior knowledge (through the use of tabular data). Using a spherical latent space to encode prototypes, we can achieve few-shot fine-grain classification by maximally separating the classes while incorporating domain knowledge as informative priors. We describe how to construct a hypersphere of prototypes that embed a-priori domain information, and demonstrate the effectiveness of the approach on challenging benchmark datasets for fine-grain classification, with top results for one-shot classification and 5x speedups in training time.