datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
SparseTable_Bench_Datasetcephalometry-novel-sparse-duplicates-single-image
Cephalometry Duplicates (Single Image Rows)
Generated from flaitenberger/cephalometry-novel-sparse-duplicates.
Generated: 2026-02-11T11:37:26.322342Z
Rows: 60
One image per row (image) with two annotation sets (points_1, points_2).
image_document_name is the representative image document.
annotation_1_document_name and annotation_2_document_name identify the two source annotation docs.
Important:
This is a structural reformat to one-image-per-row. It does not prove physical pixel… See the full description on the dataset page: https://huggingface.co/datasets/flaitenberger/cephalometry-novel-sparse-duplicates-single-image.cephalometry-novel-sparse-duplicates-grouped
Cephalometry Duplicates (Grouped, One Image Per Row)
Generated from flaitenberger/cephalometry-novel-sparse + local DB grouping.
Generated: 2026-02-11T11:39:41.052613Z
Rows: 73
One representative image per row (image)
Variable number of annotation sets per row (annotation_points_list) with annotation_set_count >= 2
Grouping rule:
group by (patient_id, created) in DB for XHeadLat docs with point results.
keep groups where at least 2 docs are present in source HF dataset and point… See the full description on the dataset page: https://huggingface.co/datasets/flaitenberger/cephalometry-novel-sparse-duplicates-grouped.