Self-supervised 3D Shape and Viewpoint Estimation from Single Images for Robotics
This addresses the problem of 3D perception for robotics by enabling practical applications with single-view data, though it is incremental as it builds on existing self-supervised methods.
The paper tackles 3D shape and viewpoint estimation from single images using self-supervision from silhouettes, eliminating the need for ground truth data or multiple views, and demonstrates improved grasping performance in robotics simulations and with a PR2 robot.
We present a convolutional neural network for joint 3D shape prediction and viewpoint estimation from a single input image. During training, our network gets the learning signal from a silhouette of an object in the input image - a form of self-supervision. It does not require ground truth data for 3D shapes and the viewpoints. Because it relies on such a weak form of supervision, our approach can easily be applied to real-world data. We demonstrate that our method produces reasonable qualitative and quantitative results on natural images for both shape estimation and viewpoint prediction. Unlike previous approaches, our method does not require multiple views of the same object instance in the dataset, which significantly expands the applicability in practical robotics scenarios. We showcase it by using the hallucinated shapes to improve the performance on the task of grasping real-world objects both in simulation and with a PR2 robot.