arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1802.07461cs.LGcs.AIcs.CVstat.ML
收藏

基于好奇心的内在动机引发的结构化行为

Emergence of Structured Behaviors from Curiosity-Based Intrinsic Motivation

Stanford University(斯坦福大学)

Nick Haber, Damian Mrowca, Li Fei-Fei, Daniel L. K. Yamins

摘要:

婴儿是玩耍的专家,在缺乏明确外在奖励信号的非结构化环境中,他们有着惊人的能力产生新颖的结构化行为。我们试图用一个神经网络来复制其中的一些能力,这个神经网络实现了好奇心驱动的内在动机。使用一个简单但生态自然主义的模拟环境,智能体可以在其中移动并与它看到的对象交互,智能体学习一个世界模型,预测其行为的动态后果。同时,智能体学会采取行动,对抗发展中的世界模型,推动智能体探索与环境的新颖和信息交互。我们证明,这一政策导致自我监督出现的一系列复杂的行为,包括自我运动预测,对象的注意力,和对象收集。此外,代理学习的世界模型支持对象动态预测和定位任务的改进性能。我们的研究结果是一个原则的证明,计算模型的内在动机可能占的关键特征,发展性视觉学习的婴儿。

英文摘要:

Infants are experts at playing, with an amazing ability to generate novel structured behaviors in unstructured environments that lack clear extrinsic reward signals. We seek to replicate some of these abilities with a neural network that implements curiosity-driven intrinsic motivation. Using a simple but ecologically naturalistic simulated environment in which the agent can move and interact with objects it sees, the agent learns a world model predicting the dynamic consequences of its actions. Simultaneously, the agent learns to take actions that adversarially challenge the developing world model, pushing the agent to explore novel and informative interactions with its environment. We demonstrate that this policy leads to the self-supervised emergence of a spectrum of complex behaviors, including ego motion prediction, object attention, and object gathering. Moreover, the world model that the agent learns supports improved performance on object dynamics prediction and localization tasks. Our results are a proof-of-principle that computational models of intrinsic motivation might account for key features of developmental visuomotor learning in infants.

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。

↑