Yale-CMU-Berkeley dataset for robotic manipulation research

TLDR

The paper introduces the Yale‑CMU‑Berkeley dataset, a collection of real‑life objects intended for benchmarking robotic manipulation research. The dataset comprises 600 high‑resolution RGB and RGB‑D images, five textured 3‑D model sets, segmentation masks, and calibration data for each object, captured with the BigBIRD scanning rig and Google scanners, and is accompanied by Python scripts and a ROS node for downloading, point‑cloud generation, and URDF creation. A dedicated website, www.ycbbenchmarks.org, hosts the dataset and facilitates publication of test results, discussion, and the development of task protocols and benchmarks.

Abstract

In this paper, we present an image and model dataset of the real-life objects from the Yale-CMU-Berkeley Object Set, which is specifically designed for benchmarking in manipulation research. For each object, the dataset presents 600 high-resolution RGB images, 600 RGB-D images and five sets of textured three-dimensional geometric models. Segmentation masks and calibration information for each image are also provided. These data are acquired using the BigBIRD Object Scanning Rig and Google Scanners. Together with the dataset, Python scripts and a Robot Operating System node are provided to download the data, generate point clouds and create Unified Robot Description Files. The dataset is also supported by our website, www.ycbbenchmarks.org , which serves as a portal for publishing and discussing test results along with proposing task protocols and benchmarks.