arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.02142cs.LGcs.AI
收藏

具有可微增强层的深度不变网络

Deep invariant networks with differentiable augmentation layers

Université Paris-Saclay(巴黎萨克雷大学);Inria(法国国家信息与自动化研究所);CEA(法国原子能和替代能源委员会)

Cédric Rommel, Thomas Moreau, Alexandre Gramfort

摘要:

在机器学习中,设计对某些数据变换不变性的学习系统是至关重要的。从业者通常可以通过选择网络架构,例如使用卷积进行翻译,或使用数据增强,在训练模型上强制执行所需的不变性。然而,在网络中实施真正的不变性可能是困难的,并且数据不变性并不总是已知的。用于学习数据增强策略的最新方法需要保留的数据并且基于双层优化问题,该双层优化问题解决起来很复杂并且通常需要计算。在这项工作中,我们研究新的方法学习不变性仅从训练数据。使用直接在网络中构建的可学习增强层,我们证明了我们的方法是非常通用的。它可以包含任何类型的可微增强,并应用于计算机视觉之外的广泛的学习问题。我们提供的经验证据表明,我们的方法是更容易和更快的训练比现代自动数据增强技术的基础上双层优化,同时实现可比的结果。实验表明,虽然通过自动数据增强转移到模型的不变性受到模型表达能力的限制,但我们的方法产生的不变性对它不敏感。

英文摘要:

Designing learning systems which are invariant to certain data transformations is critical in machine learning. Practitioners can typically enforce a desired invariance on the trained model through the choice of a network architecture, e.g. using convolutions for translations, or using data augmentation. Yet, enforcing true invariance in the network can be difficult, and data invariances are not always known a piori. State-of-the-art methods for learning data augmentation policies require held-out data and are based on bilevel optimization problems, which are complex to solve and often computationally demanding. In this work we investigate new ways of learning invariances only from the training data. Using learnable augmentation layers built directly in the network, we demonstrate that our method is very versatile. It can incorporate any type of differentiable augmentation and be applied to a broad class of learning problems beyond computer vision. We provide empirical evidence showing that our approach is easier and faster to train than modern automatic data augmentation techniques based on bilevel optimization, while achieving comparable results. Experiments show that while the invariances transferred to a model through automatic data augmentation are limited by the model expressivity, the invariance yielded by our approach is insensitive to it by design.

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。

↑