DynaHOI: Benchmarking Hand-Object Interaction for Dynamic Target
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911443003113472 |
|---|---|
| author | Hu, BoCheng Zhao, Zhonghan Zhou, Kaiyue Wang, Hongwei Wang, Gaoang |
| author_facet | Hu, BoCheng Zhao, Zhonghan Zhou, Kaiyue Wang, Hongwei Wang, Gaoang |
| contents | Most existing hand motion generation benchmarks for hand-object interaction (HOI) focus on static objects, leaving dynamic scenarios with moving targets and time-critical coordination largely untested. To address this gap, we introduce the DynaHOI-Gym, a unified online closed-loop platform with parameterized motion generators and rollout-based metrics for dynamic capture evaluation. Built on DynaHOI-Gym, we release DynaHOI-10M, a large-scale benchmark with 10M frames and 180K hand capture trajectories, whose target motions are organized into 8 major categories and 22 fine-grained subcategories. We also provide a simple observe-before-act baseline (ObAct) that integrates short-term observations with the current frame via spatiotemporal attention to predict actions, achieving an 8.1% improvement in location success rate. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_11919 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DynaHOI: Benchmarking Hand-Object Interaction for Dynamic Target Hu, BoCheng Zhao, Zhonghan Zhou, Kaiyue Wang, Hongwei Wang, Gaoang Computer Vision and Pattern Recognition Artificial Intelligence Most existing hand motion generation benchmarks for hand-object interaction (HOI) focus on static objects, leaving dynamic scenarios with moving targets and time-critical coordination largely untested. To address this gap, we introduce the DynaHOI-Gym, a unified online closed-loop platform with parameterized motion generators and rollout-based metrics for dynamic capture evaluation. Built on DynaHOI-Gym, we release DynaHOI-10M, a large-scale benchmark with 10M frames and 180K hand capture trajectories, whose target motions are organized into 8 major categories and 22 fine-grained subcategories. We also provide a simple observe-before-act baseline (ObAct) that integrates short-term observations with the current frame via spatiotemporal attention to predict actions, achieving an 8.1% improvement in location success rate. |
| title | DynaHOI: Benchmarking Hand-Object Interaction for Dynamic Target |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2602.11919 |