arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.11702stat.MLcs.LG
收藏

随机原型嵌入

Stochastic Prototype Embeddings

University of Colorado, Boulder(科罗拉多大学博尔德分校);Sensory Inc.(感官公司);Google Research(谷歌研究院)

Tyler R. Scott, Karl Ridgeway, Michael C. Mozer

摘要:

监督深度嵌入方法将域的输入投影到表示空间,在该表示空间中,同类实例彼此靠近,而不同类实例相距较远。我们提出了一种概率方法,将嵌入作为随机变量。扩展最先进的确定性方法,原型网络(Snell等人,2017),我们的方法假设存在一个类原型,类实例围绕该原型高斯分布。原型后验是一个产品的分布在标记的实例,和查询实例分类边缘化相对原型接近嵌入不确定性。我们描述了一个有效的采样器近似推理,使我们能够训练模型在大致相同的空间和时间成本,其确定性的兄弟。消除不确定性可以提高Few-Shot学习的性能,并优雅地处理标签噪声和分布外输入。与现有技术的随机方法相比,Hedged Instance Embeddings(Oh等人,2019年),我们实现了卓越的大和开集分类精度。我们的方法还将类别区分特征与嵌入空间的轴对齐,产生可解释的、解纠缠的表示。

英文摘要:

Supervised deep-embedding methods project inputs of a domain to a representational space in which same-class instances lie near one another and different-class instances lie far apart. We propose a probabilistic method that treats embeddings as random variables. Extending a state-of-the-art deterministic method, Prototypical Networks (Snell et al., 2017), our approach supposes the existence of a class prototype around which class instances are Gaussian distributed. The prototype posterior is a product distribution over labeled instances, and query instances are classified by marginalizing relative prototype proximity over embedding uncertainty. We describe an efficient sampler for approximate inference that allows us to train the model at roughly the same space and time cost as its deterministic sibling. Incorporating uncertainty improves performance on few-shot learning and gracefully handles label noise and out-of-distribution inputs. Compared to the state-of-the-art stochastic method, Hedged Instance Embeddings (Oh et al., 2019), we achieve superior large- and open-set classification accuracy. Our method also aligns class-discriminating features with the axes of the embedding space, yielding an interpretable, disentangled representation.

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。

↑