HOVA-500K

From the catalogue. Listed by one researcher from one source page.

• id: hova-500k
• kind: language and reasoning annotation
• robot or device: human behaviour images with affordance points (class: sensor only)
• how collected: affordance annotation from human behaviours
• size as stated: 500,000 images; 1,726 object categories; 675 actions
• hours: not stated | hours only claimed: none | episodes: not stated
• year: 2025
• organisation: JiaaZ (HF account)
• licence: not stated (class: not stated)
• access: open (class: open)
• link (dataset): https://huggingface.co/datasets/JiaaZ/HOVA-500K
• Hugging Face repository: JiaaZ/HOVA-500K, 214 downloads in 30 days, https://huggingface.co/datasets/JiaaZ/HOVA-500K
• note: From GLOVER++; no card on the Hub.

Confirm the licence at the link before relying on it.