Learning to Poke by Poking: Experiential Learning of Intuitive Physics
Learning to Poke by Poking: Experiential Learning of Intuitive Physics
复制标题
DOI：
--
复制
发表时间：
2016-06
期刊：
影响因子：
--
通讯作者：
Pulkit Agrawal;Ashvin Nair;P. Abbeel;Jitendra Malik;S. Levine
中科院分区：
文献类型：
--
作者：
Pulkit Agrawal;Ashvin Nair;P. Abbeel;Jitendra Malik;S. Levine
我们调查的经验学习范式获得直观物理学的内部模型。我们的模型进行评估，在现实世界中的机器人操作任务，需要通过戳位移对象到目标位置。该机器人通过在不同物体上执行超过10万次戳击，积累了超过400小时的经验。我们提出了一种基于深度神经网络的新方法，通过联合估计动力学的正向和反向模型，直接从图像中建模机器人交互的动力学。逆模型目标提供监督以构造信息丰富的视觉特征，然后正模型可以预测这些视觉特征，并进而正则化逆模型的特征空间。这两个目标之间的相互作用创建了有用的，准确的模型，然后可以用于多步决策。该公式具有额外的好处，即可以在抽象特征空间中学习前向模型，从而减轻预测像素的需要。我们的实验表明，这种联合建模方法优于其他方法。
We investigate an experiential learning paradigm for acquiring an internal model of intuitive physics. Our model is evaluated on a real-world robotic manipulation task that requires displacing objects to target locations by poking. The robot gathered over 400 hours of experience by executing more than 100K pokes on different objects. We propose a novel approach based on deep neural networks for modeling the dynamics of robot's interactions directly from images, by jointly estimating forward and inverse models of dynamics. The inverse model objective provides supervision to construct informative visual features, which the forward model can then predict and in turn regularize the feature space for the inverse model. The interplay between these two objectives creates useful, accurate models that can then be used for multi-step decision making. This formulation has the additional benefit that it is possible to learn forward models in an abstract feature space and thus alleviate the need of predicting pixels. Our experiments show that this joint modeling approach outperforms alternative methods.