Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

YCB-Handovers Dataset: Analyzing Object Weight Impact on Human Handovers to Adapt Robotic Handover Motion

Parag Khanna, Karen Jane Dsouza, Chunyu Wang, Mårten Björkman, Christian Smith

TL;DR

The paper introduces the YCB-Handovers dataset, a motion-capture resource of 2771 human-human handovers using the YCB object set to study how object weight affects handover motion and readiness cues for robotic planning. It provides a full methodological pipeline from experimental design to data post-processing, and performs extensive analyses showing weight influences kinematics and that weight-based classification can be highly accurate with motion data. The work demonstrates both unsupervised and supervised approaches for weight classification, and reveals that weight is a dominant factor in handover dynamics, with carefulness playing a secondary role, particularly as objects become heavier. The dataset offers a transferable benchmark for developing weight-sensitive, human-inspired robotic handover strategies and motivates future work to broaden participant diversity and real-world testing in industrial and assistive robotics.

Abstract

This paper introduces the YCB-Handovers dataset, capturing motion data of 2771 human-human handovers with varying object weights. The dataset aims to bridge a gap in human-robot collaboration research, providing insights into the impact of object weight in human handovers and readiness cues for intuitive robotic motion planning. The underlying dataset for object recognition and tracking is the YCB (Yale-CMU-Berkeley) dataset, which is an established standard dataset used in algorithms for robotic manipulation, including grasping and carrying objects. The YCB-Handovers dataset incorporates human motion patterns in handovers, making it applicable for data-driven, human-inspired models aimed at weight-sensitive motion planning and adaptive robotic behaviors. This dataset covers an extensive range of weights, allowing for a more robust study of handover behavior and weight variation. Some objects also require careful handovers, highlighting contrasts with standard handovers. We also provide a detailed analysis of the object's weight impact on the human reaching motion in these handovers.

YCB-Handovers Dataset: Analyzing Object Weight Impact on Human Handovers to Adapt Robotic Handover Motion

TL;DR

The paper introduces the YCB-Handovers dataset, a motion-capture resource of 2771 human-human handovers using the YCB object set to study how object weight affects handover motion and readiness cues for robotic planning. It provides a full methodological pipeline from experimental design to data post-processing, and performs extensive analyses showing weight influences kinematics and that weight-based classification can be highly accurate with motion data. The work demonstrates both unsupervised and supervised approaches for weight classification, and reveals that weight is a dominant factor in handover dynamics, with carefulness playing a secondary role, particularly as objects become heavier. The dataset offers a transferable benchmark for developing weight-sensitive, human-inspired robotic handover strategies and motivates future work to broaden participant diversity and real-world testing in industrial and assistive robotics.

Abstract

This paper introduces the YCB-Handovers dataset, capturing motion data of 2771 human-human handovers with varying object weights. The dataset aims to bridge a gap in human-robot collaboration research, providing insights into the impact of object weight in human handovers and readiness cues for intuitive robotic motion planning. The underlying dataset for object recognition and tracking is the YCB (Yale-CMU-Berkeley) dataset, which is an established standard dataset used in algorithms for robotic manipulation, including grasping and carrying objects. The YCB-Handovers dataset incorporates human motion patterns in handovers, making it applicable for data-driven, human-inspired models aimed at weight-sensitive motion planning and adaptive robotic behaviors. This dataset covers an extensive range of weights, allowing for a more robust study of handover behavior and weight variation. Some objects also require careful handovers, highlighting contrasts with standard handovers. We also provide a detailed analysis of the object's weight impact on the human reaching motion in these handovers.
Paper Structure (21 sections, 5 figures, 13 tables)

This paper contains 21 sections, 5 figures, 13 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methods
  4. Experimental Design
  5. Data Collection
  6. Data Post-Processing
  7. Dataset Description
  8. Analysis: Impact of Object Weight on motion
  9. Velocities and Acceleration
  10. Weight Groups
  11. Clustering for Weight Classification
  12. Unsupervised Learning: K-Means Clustering
  13. Supervised Learning: KNN, SVM, and Random Forest
  14. Carefulness and Weight
  15. Classification Accuracy for Four Labels
  16. ...and 6 more sections

Figures (5)

  • Figure 1: YCB-Handovers Dataset: This dataset captures handover interactions involving various objects from the YCB object dataset. The handovers were recorded in a Motion Capture room to ensure accurate tracking of movements. The left image illustrates a handover of a marker pen between two participants, while the right image showcases various objects used in these recorded handovers.
  • Figure 2: Different objects used in the the different baskets
  • Figure 3: The image shows the position of the dominant hand during two consecutive handovers of Pair 1 for the non-careful trial with the YCB pitcher.
  • Figure 4: Motion characteristics for human hand motion with different weighted objects in human-human handovers. Mean values observed for a particular object are plotted. Note: This plot excluddes the objects with added carefullness
  • Figure 5: Motion Characteristics across different Weight Categories