Vigen1/hypersim-mini
Hypersim Minimal (RGB + Semantic Mapped) This dataset is a minimal extraction from Hypersim: RGB: preview (either preview JPGs or HDR color.hdf5 tonemapped to PNG) Semantic labels: mapped to uint8 PNG using clip40to39 Columns image — RGB image mask — segmentation mask (uint8) scene, camera, frame — identifiers Splits train validation (ratio: 0.1) Note: You are responsible for complying with the original Hypersim license/terms.
055
Hypersim Minimal (RGB + Semantic Mapped)
This dataset is a minimal extraction from Hypersim:
- RGB:
preview(either preview JPGs or HDR color.hdf5 tonemapped to PNG) - Semantic labels: mapped to uint8 PNG using
clip40to39
Columns
image— RGB imagemask— segmentation mask (uint8)scene,camera,frame— identifiers
Splits
trainvalidation(ratio: 0.1)
Note: You are responsible for complying with the original Hypersim license/terms.
