dataset · Global

FSCM_Flood_nerf

Hugging Face dataset ParkSY/FSCM_Flood_nerf

Claim this listing

Represent FSCM_Flood_nerf? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Data offered docs-verified: flood image-edit pairs (input_image and edited_image) · edit_prompt text describing the flood change · label (int64 column) · depthmap column (float64 type) · normalmap column (float64 type) · train split of 30,285 rows
Formats docs-verified: parquet
Access llm-inferred: Public Hugging Face dataset; licence not stated (ENTRY licence is null)

Use cases llm-inferred: Training or evaluating text-guided flood-intensity image edit models from input/edited image pairs · Generating illustrative flood-extent imagery for client or public flood-risk visualisation, subject to licence checks · Prototyping depth- and normal-map conditioned flood scene edits for 3D or NeRF-style scene work
Audience llm-inferred: computer vision and ML engineers · flood risk visualisation specialists · planners (indirectly, via scenario imagery)

Constraints llm-inferred: Docs give no geographic or temporal coverage; ENTRY region 'Global' is a catalogue tag, not documented coverage · Previewed prompts all concern intensifying floodwater; wider scenario diversity is not shown · depthmap and normalmap are null in every previewed row; meaning of label is undocumented · README.md is empty: no dataset card, provenance (real vs generated images) or licence · Large download: 41 GB total for about 30k rows

Verified 2026-10-15 via HF API: exists; 275 downloads/30d, 0 likes.

Metadata observed 14/10/2026 · research sweep: manual-2026-10-15 (Hugging Face datasets)