Jacquard: A Large Scale Dataset for Robotic Grasp Detection
This provides a scalable dataset for robotic grasping research, addressing a key bottleneck in training deep networks for real-life robotisation applications.
The authors tackled the lack of labeled data for robotic grasp detection by creating Jacquard, a large-scale synthetic dataset based on ShapeNet, which includes RGB-D images and annotations from simulated grasp attempts. Experiments showed that Jacquard enables much better generalization than human-labeled datasets, as validated by real robot trials.
Grasping skill is a major ability that a wide number of real-life applications require for robotisation. State-of-the-art robotic grasping methods perform prediction of object grasp locations based on deep neural networks. However, such networks require huge amount of labeled data for training making this approach often impracticable in robotics. In this paper, we propose a method to generate a large scale synthetic dataset with ground truth, which we refer to as the Jacquard grasping dataset. Jacquard is built on a subset of ShapeNet, a large CAD models dataset, and contains both RGB-D images and annotations of successful grasping positions based on grasp attempts performed in a simulated environment. We carried out experiments using an off-the-shelf CNN, with three different evaluation metrics, including real grasping robot trials. The results show that Jacquard enables much better generalization skills than a human labeled dataset thanks to its diversity of objects and grasping positions. For the purpose of reproducible research in robotics, we are releasing along with the Jacquard dataset a web interface for researchers to evaluate the successfulness of their grasping position detections using our dataset.