Scalable Uncertainty for Computer Vision With Functional Variational Inference
Scalable Uncertainty for Computer Vision With Functional Variational Inference
复制标题
DOI：
10.1109/cvpr42600.2020.01202
复制
发表时间：
2020-03
期刊：
影响因子：
--
通讯作者：
Eduardo D C Carvalho;R. Clark;Andrea Nicastro;P. Kelly
中科院分区：
文献类型：
--
作者：
Eduardo D C Carvalho;R. Clark;Andrea Nicastro;P. Kelly
随着深度学习继续在计算机视觉领域取得成功应用，量化所有形式的不确定性的能力是其在现实世界中安全可靠部署的最高要求。在这项工作中，我们利用函数空间中的变分推理公式，其中我们将高斯过程(GP)与贝叶斯CNN先验和变分族相关联。由于GP完全由它们的均值和协方差函数决定，因此我们能够以通过任何选择的CNN架构和任何监督学习任务的单次前向传递的代价来获得预测不确定性估计。通过利用诱导协方差矩阵的结构，我们提出了数值上有效的算法，使得在高维任务(如深度估计和语义分割)中的快速训练成为可能。此外，我们还给出了构造概率对应的回归损失函数与任意不确定性量化相容的充分条件。
As Deep Learning continues to yield successful applications in Computer Vision, the ability to quantify all forms of uncertainty is a paramount requirement for its safe and reliable deployment in the real-world. In this work, we leverage the formulation of variational inference in function space, where we associate Gaussian Processes (GPs) to both Bayesian CNN priors and variational family. Since GPs are fully determined by their mean and covariance functions, we are able to obtain predictive uncertainty estimates at the cost of a single forward pass through any chosen CNN architecture and for any supervised learning task. By leveraging the structure of the induced covariance matrices, we propose numerically efficient algorithms which enable fast training in the context of high-dimensional tasks such as depth estimation and semantic segmentation. Additionally, we provide sufficient conditions for constructing regression loss functions whose probabilistic counterparts are compatible with aleatoric uncertainty quantification.