arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1712.09708cs.CVcs.LG
收藏

学习更多的通用表示用于迁移学习

Learning More Universal Representations for Transfer-Learning

CEA, LIST(法国原子能和替代能源委员会LIST研究所)

Youssef Tamaazousti, Hervé Le Borgne, Céline Hudelot, Mohamed El Amine Seddik, Mohamed Tamaazousti

摘要:

如果一个表示对视觉世界的任何元素进行编码,那么它被认为是通用的(例如,对象,场景)以任何配置(例如,规模、背景)。虽然不期望纯粹的普遍表示,但文献中的目标是从具有一定水平的表示开始提高普遍性水平。要做到这一点,最先进的技术在于学习多样化训练问题上的基于CNN的表示(例如,ImageNet通过添加注释数据进行了修改)。虽然它有效地增加了通用性,这种方法仍然需要大量的努力,以满足注释数据的需求。在这项工作中,我们提出了两种方法来提高通用性,但要特别注意限制注释数据的需要。我们还提出了一个统一的框架的基础上多样化的训练问题的方法。最后,为了更好地匹配阿特金森关于普遍人类表征的认知研究,我们提出了依赖于迁移学习方案以及一个新的度量来评估普遍性。后者的目的是展示我们的方法对10个目标问题的兴趣,这些问题与分类任务和各种视觉领域有关。

英文摘要:

A representation is supposed universal if it encodes any element of the visual world (e.g., objects, scenes) in any configuration (e.g., scale, context). While not expecting pure universal representations, the goal in the literature is to improve the universality level, starting from a representation with a certain level. To do so, the state-of-the-art consists in learning CNN-based representations on a diversified training problem (e.g., ImageNet modified by adding annotated data). While it effectively increases universality, such approach still requires a large amount of efforts to satisfy the needs in annotated data. In this work, we propose two methods to improve universality, but pay special attention to limit the need of annotated data. We also propose a unified framework of the methods based on the diversifying of the training problem. Finally, to better match Atkinson's cognitive study about universal human representations, we proposed to rely on the transfer-learning scheme as well as a new metric to evaluate universality. This latter, aims us to demonstrates the interest of our methods on 10 target-problems, relating to the classification task and a variety of visual domains.

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。

↑