arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2108.08891cs.CVcs.LG
收藏

神经TMD层:通过SDE生成器对特征的瞬时流进行建模

Neural TMDlayer: Modeling Instantaneous flow of features via SDE Generators

University of Wisconsin-Madison(威斯康星大学麦迪逊分校);University of Illinois at Chicago(伊利诺伊大学芝加哥分校)

Zihang Meng, Vikas Singh, Sathya N. Ravi

摘要:

我们研究了如何随机微分方程(SDG)为基础的想法可以激发新的修改现有的算法在计算机视觉的一组问题。宽松地说,我们的配方是有关的显式和隐式的策略数据增强和组等方差,但来自新的结果估计无穷小发电机的一类随机过程的文献。如果并且当应用程序/任务的需求与我们可以有效处理的过程类型的固有属性和行为之间存在名义上的协议时,我们可以获得一个非常简单和有效的插件层,该插件层可以并入任何现有的网络架构中,只需最小的修改和一些额外的参数。我们展示了一些有前途的视觉任务,包括少样本学习,点云Transformers和深度变分分割获得效率或性能改进的实验。

英文摘要:

We study how stochastic differential equation (SDE) based ideas can inspire new modifications to existing algorithms for a set of problems in computer vision. Loosely speaking, our formulation is related to both explicit and implicit strategies for data augmentation and group equivariance, but is derived from new results in the SDE literature on estimating infinitesimal generators of a class of stochastic processes. If and when there is nominal agreement between the needs of an application/task and the inherent properties and behavior of the types of processes that we can efficiently handle, we obtain a very simple and efficient plug-in layer that can be incorporated within any existing network architecture, with minimal modification and only a few additional parameters. We show promising experiments on a number of vision tasks including few shot learning, point cloud transformers and deep variational segmentation obtaining efficiency or performance improvements.

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。

↑