Competition

The competition will consist of both functionality and task benchmarks. The following benchmarks will be (tentatively) conducted during this field campaign.

Functionality Benchmarks (FBM):

  • Object Detection

  • Person Detection

  • Gesture Recognition

  • Activity Recognition

  • Robot-to-Human Handover

  • Human-to-Robot Handover

Task Benchmarks (TBM):

  • Assess Activity State

  • Item Delivery

Rulebook

More details about each benchmark, including the procedure and evaluation metrics can be found in the rulebook:

https://github.com/HEART-MET/rulebook

Schedule