TACO: Benchmarking Generalizable Bimanual Tool-ACtion-Object Understanding
This addresses the problem of limited data for analyzing and synthesizing bimanual tool manipulations in robotics and AI, though it is incremental as it builds on existing hand-object interaction research.
The authors tackled the lack of data for bimanual hand-object interactions by constructing TACO, a dataset with 2.5K motion sequences, and used it to benchmark tasks like action recognition and motion forecasting, revealing new challenges and opportunities.
Humans commonly work with multiple objects in daily life and can intuitively transfer manipulation skills to novel objects by understanding object functional regularities. However, existing technical approaches for analyzing and synthesizing hand-object manipulation are mostly limited to handling a single hand and object due to the lack of data support. To address this, we construct TACO, an extensive bimanual hand-object-interaction dataset spanning a large variety of tool-action-object compositions for daily human activities. TACO contains 2.5K motion sequences paired with third-person and egocentric views, precise hand-object 3D meshes, and action labels. To rapidly expand the data scale, we present a fully automatic data acquisition pipeline combining multi-view sensing with an optical motion capture system. With the vast research fields provided by TACO, we benchmark three generalizable hand-object-interaction tasks: compositional action recognition, generalizable hand-object motion forecasting, and cooperative grasp synthesis. Extensive experiments reveal new insights, challenges, and opportunities for advancing the studies of generalizable hand-object motion analysis and synthesis. Our data and code are available at https://taco2024.github.io.