arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.13356cs.CVcs.LG
收藏

基于李群的鲁棒自监督学习

Robust Self-Supervised Learning with Lie Groups

Fundamental AI Research (FAIR), Meta AI(Meta AI 基础人工智能研究院(FAIR))

Mark Ibrahim, Diane Bouchacourt, Ari Morcos

摘要:

深度学习在计算机视觉领域取得了显著的进步。即便如此,今天最好的模型在与训练过程中看到的变化略有不同时也是脆弱的。物体姿态、颜色或照明的微小变化可能导致灾难性的错误分类。最先进的模型很难理解一组变化如何影响不同的对象。我们提出了一个框架,灌输的概念,对象如何在更现实的设置不同。我们的方法应用李群的形式主义来捕获连续变换,以提高模型对分布变化的鲁棒性。我们将我们的框架应用于最先进的自监督学习(SSL)模型之上,发现用李群明确建模变换会导致MAE在两个已知实例上的性能增益大于10%,这两个已知实例在典型姿势中出现在新姿势中,以及在任何姿势中的未知实例上。我们还将我们的方法应用于ImageNet,发现Lie算子将性能提高了近4%。这些结果证明了学习变换提高模型鲁棒性的前景。

英文摘要:

Deep learning has led to remarkable advances in computer vision. Even so, today's best models are brittle when presented with variations that differ even slightly from those seen during training. Minor shifts in the pose, color, or illumination of an object can lead to catastrophic misclassifications. State-of-the art models struggle to understand how a set of variations can affect different objects. We propose a framework for instilling a notion of how objects vary in more realistic settings. Our approach applies the formalism of Lie groups to capture continuous transformations to improve models' robustness to distributional shifts. We apply our framework on top of state-of-the-art self-supervised learning (SSL) models, finding that explicitly modeling transformations with Lie groups leads to substantial performance gains of greater than 10% for MAE on both known instances seen in typical poses now presented in new poses, and on unknown instances in any pose. We also apply our approach to ImageNet, finding that the Lie operator improves performance by almost 4%. These results demonstrate the promise of learning transformations to improve model robustness.

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。

↑