我的房子,我的规则:用图神经网络学习整理偏好
My House, My Rules: Learning Tidying Preferences with Graph Neural Networks
摘要:
布置家居物品的机器人应该根据用户的偏好来做,这些偏好本质上是主观的,难以建模。我们呈现NeatNet:一种新的变分自编码器架构,使用图神经网络层,它可以通过观察用户如何安排场景来提取低维潜在偏好向量。给定任何一组对象,该向量然后可以用于生成针对该用户的空间偏好定制的布置,其中词嵌入用于概括到新对象。我们开发了一个整理模拟器收集重排的例子,从75个用户,并证明经验表明,我们的方法始终产生整洁和个性化的安排在各种重排方案。
英文摘要:
Robots that arrange household objects should do so according to the user's preferences, which are inherently subjective and difficult to model. We present NeatNet: a novel Variational Autoencoder architecture using Graph Neural Network layers, which can extract a low-dimensional latent preference vector from a user by observing how they arrange scenes. Given any set of objects, this vector can then be used to generate an arrangement which is tailored to that user's spatial preferences, with word embeddings used for generalisation to new objects. We develop a tidying simulator to gather rearrangement examples from 75 users, and demonstrate empirically that our method consistently produces neat and personalised arrangements across a variety of rearrangement scenarios.
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。