Abstract

Abstract. We propose a framework for automatic modeling, detection, and tracking of 3D objects with a Kinect. The detection part is mainly based on the recent template-based LINEMOD approach [1] for object detection. We show how to build the templates automatically from 3D models, and how to estimate the 6 degrees-of-freedom pose accurately and in real-time. The pose estimation and the color information allow us to check the detection hypotheses and improves the correct detection rate by 13% with respect to the original LINEMOD. These many improvements make our framework suitable for object manipulation in Robotics applications. Moreover we propose a new dataset made of 15 registered, 1100+ frame video sequences of 15 various objects for the evaluation of future competing methods. Fig. 1. 15 different texture-less 3D objects are simultaneously detected with our approach under different poses on heavy cluttered background with partial occlusion. Each detected object is augmented with its 3D model. We also show the corresponding coordinate systems.

No abstract

Instrumenting and collecting annotated visual grasping datasets to train modern machine learning algorithms can be extremely time-consuming and expensive. An appealing alternative is to use off-the-shelf simulators to render synthetic data for which ground-truth annotations are generated automatically. Unfortunately, models trained purely on simulated data often fail to generalize to the real world. We study how randomized simulated environments and domain adaptation methods can be extended to train a grasping system to grasp novel objects from raw monocular RGB images. We extensively evaluate our approaches with a total of more than 25,000 physical test grasps, studying a range of simulation conditions and domain adaptation methods, including a novel extension of pixel-level domain adaptation that we term the GraspGAN. We show that, by using synthetic data and domain adaptation, we are able to reduce the number of real-world samples needed to achieve a given level of performance by up to 50 times, using only randomly generated simulated objects. We also show that by using only unlabeled real-world data and our GraspGAN methodology, we obtain real-world grasping performance without any real-world labels that is similar to that achieved with 939,777 labeled real-world samples.

No abstract