en un clic
OpenH-RF
OpenH-RF contient 2 skills collectées depuis open-h, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Evaluate a submission to the OpenH-RF (open ultrasound channel-data) initiative against the RFP and submission guide. Use this whenever the user wants to review, grade, validate, accept, or QA a proposed contribution to OpenH-RF, OpenH, or any open ultrasound channel/RF dataset — including when they hand over a folder, a Hugging Face link, a zea file, or a data card and ask whether it's ready, whether it meets the bar, what's missing, or whether it can be accepted. Also use when the user wants to auto-fix gaps in a submission, such as generating a missing B-mode pipeline, filling in data-card fields, or rendering a reference reconstruction. Trigger even if the user doesn't say OpenH-RF by name — ultrasound channel data plus a review or intake context is enough.
Convert a source ultrasound dataset (HDF5, Verasonics .mat, .npz, raw .bin + .json sidecars, a Hugging Face dataset) into the OpenH-RF (zea) file format. Inspects a sample, proposes a source->zea field mapping for the contributor to confirm, then generates and verifies convert.py + a reconstruction against the openh-rf-latest spec. Does NOT write the data card.