DexYCB: A Benchmark for Capturing Hand Grasping of Objects
This provides a benchmark for researchers in robotics and computer vision to evaluate hand-object interaction methods, but it is incremental as it builds on existing datasets and tasks.
The authors introduced DexYCB, a new dataset for capturing hand grasping of objects, and benchmarked state-of-the-art methods on tasks like 2D detection, 6D pose estimation, and 3D hand pose estimation, with results showing competitive performance and application to safe robot grasps in human-to-robot handovers.
We introduce DexYCB, a new dataset for capturing hand grasping of objects. We first compare DexYCB with a related one through cross-dataset evaluation. We then present a thorough benchmark of state-of-the-art approaches on three relevant tasks: 2D object and keypoint detection, 6D object pose estimation, and 3D hand pose estimation. Finally, we evaluate a new robotics-relevant task: generating safe robot grasps in human-to-robot object handover. Dataset and code are available at https://dex-ycb.github.io.