BriefGPT.xyz
Apr, 2023
强健不变表示中的领域通用性
Domain Generalization In Robust Invariant Representation
HTML
PDF
Gauri Gupta, Ritvik Kapila, Keshav Gupta, Ramesh Raskar
TL;DR
本文研究对象识别中不变表示的泛化性,经过广泛实验,我们证明了不变模型学习到的非结构化潜在表示对分配偏移具有鲁棒性,因此使不变性成为有限资源环境中训练的理想属性。
Abstract
Unsupervised approaches for learning representations invariant to common transformations are used quite often for
object recognition
. Learning
invariances
makes models more robust and practical to use in real-wor
→