BriefGPT.xyz
Nov, 2023
属性多样性决定了VQA中的系统性差距
Attribute Diversity Determines the Systematicity Gap in VQA
HTML
PDF
Ian Berlot-Attwell, A. Michael Carrell, Kumar Krishna Agrawal, Yash Sharma, Naomi Saphra
TL;DR
通过研究视觉问答中的系统性差异,我们发现神经网络在泛化到新的熟悉概念组合时的表现差异。尽管训练数据的数量增加并不能减小系统性差距,但训练数据中未见组合的属性多样性可以减小系统性差距。因此,我们的实验证明,在训练过程中看到更多不同的属性类型组合,我们可以期望结果模型更具系统性。
Abstract
The degree to which
neural networks
can
generalize
to new combinations of familiar concepts, and the conditions under which they are able to do so, has long been an open question. In this work, we study the
→