arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.11450cs.CVcs.LGeess.IV
收藏

基于超球面先验知识的细粒度Few-Shot视觉

Fine-Grain Few-Shot Vision via Domain Knowledge as Hyperspherical Priors

Augustus Intelligence(奥古斯都智能公司)

Bijan Haney, Alexander Lavin

摘要:

原型网络已被证明在计算机视觉中的少样本学习任务中表现良好。然而,当类别彼此非常相似(细粒度分类)并且目前无法考虑先验知识(通过使用表格数据)时,这些网络就会陷入困境。使用一个球形的潜在空间编码的原型,我们可以实现Few-Shot细粒度分类最大限度地分离类,同时将领域知识作为信息先验。我们描述了如何构建一个嵌入先验域信息的原型超球体,并展示了该方法在挑战细粒度分类基准数据集上的有效性,最佳结果是一次性分类和5倍的训练时间加速。

英文摘要:

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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑