Datasetpublic

Shanmuk4622/cr-rvs-radar-ecg-inventory-v2

CR-RVS Radar/ECG — Dataset Inventory & Quality Census Machine-readable inventory of the clinically recorded radar vital-signs dataset of Schellenberger et al. (Scientific Data 7:291, 2020), produced as stage 1 of the CardioMamba-Net project on contactless ECG reconstruction from 24 GHz continuous-wave radar. This repo contains metadata, statistics, previews and figures — not the full raw dataset. For the raw recordings see figshare 10.6084/m9.figshare.12186516. What… See the full description on the dataset page: https://huggingface.co/datasets/Shanmuk4622/cr-rvs-radar-ecg-inventory-v2.

sourceHugging Facecc-by-4.0updated 25d agoView on Hugging Face

0likes102downloads

Dataset Card

CR-RVS Radar/ECG — Dataset Inventory & Quality Census

Machine-readable inventory of the clinically recorded radar vital-signs dataset of Schellenberger et al. (Scientific Data 7:291, 2020), produced as stage 1 of the CardioMamba-Net project on contactless ECG reconstruction from 24 GHz continuous-wave radar.

This repo contains metadata, statistics, previews and figures — not the full raw dataset. For the raw recordings see figshare 10.6084/m9.figshare.12186516.

What is here

Path	Contents
`inventory.parquet` / `inventory.csv`	One row per (subject, scenario) recording, ~60 columns
`crosscheck.csv`	Our computed durations and window counts vs. those published
`variables.json`	The `.mat` variable schema, with shapes and dtypes
`verdict.json`	Source-integrity check
`previews/*.npz`	60-second full-rate excerpt of every recording, 8 derived channels + ECG
`previews/*__bta.npy`	Beat-triggered average of radar acceleration, one per recording
`decimated/*.npz`	Whole corpus decimated to 128 Hz (8 radar channels + filtered ECG)
`figures/*.png`	Coverage heatmap, example signals, I/Q ellipse, spectra, quality panels
`report.md`	Human-readable findings
`hf_sync.py`	Resumable rate-limited HF uploader used across the project

Inventory columns

—Identity — subject, scenario, scenario_canon, timestamp, rel_path, mat_format, size_mb
—Timing — fs_radar, fs_ecg, n_radar, n_ecg, duration_s, n_samples_128hz, n_windows_nooverlap, n_windows_overlap50
—Receiver — iq_xc, iq_yc, iq_a, iq_b, iq_theta_rad, iq_gain_imbalance, iq_residual
—Per channel (I, Q, phi, dy, vel, acc, amp, cardiac, ecg1, ecg2, icg, z0, bp) — _min, _max, _mean, _std, _p01, _p99, _n_nan, _n_inf, _pct_clip, _max_flat_run
—ECG quality — ecg1_n_rpeaks, ecg1_mean_hr, ecg1_sd_hr, ecg1_rmssd_ms, ecg1_median_rr_ms, ecg1_pct_flat, ecg1_hr_plausible
—Sync & coupling — sync_lag_s, sync_xcorr, beat_coupling
—Derived — displacement_range_mm, phase_wraps, flags, n_flags, census_seconds

ecg1_rmssd_ms and ecg1_median_rr_ms are in genuine milliseconds.

Provenance

Generated by 01_verify_and_download.ipynb on Kaggle (CPU), run nb01_inventory_v2, 2026-09-02 12:48 UTC. Source: Kaggle mirror pedababugaddala/datasets-file. Verdict: RAW_MAT_TREE.

Cite the original work

bibtex

@article{schellenberger2020dataset,
  title   = {A dataset of clinically recorded radar vital signs with synchronised reference sensor signals},
  author  = {Schellenberger, Sven and Shi, Kilin and Steigleder, Tobias and Malessa, Anke and
             Michler, Fabian and Hameyer, Laura and Neumann, Nina and Lurz, Fabian and
             Ostgathe, Christoph and Weigel, Robert and Koelpin, Alexander},
  journal = {Scientific Data}, volume = {7}, number = {1}, pages = {291}, year = {2020},
  doi     = {10.1038/s41597-020-00629-5}
}

@article{chowdhury2024ecg,
  title   = {ECG waveform generation from radar signals: A deep learning perspective},
  author  = {Chowdhury, Farhana Ahmed and Hosain, Md Kamal and Islam, Md Sakib Bin and
             Hossain, Md Shafayet and Basak, Promit and Mahmud, Sakib and
             Murugappan, M. and Chowdhury, Muhammad E. H.},
  journal = {Computers in Biology and Medicine}, volume = {176}, pages = {108555}, year = {2024},
  doi     = {10.1016/j.compbiomed.2024.108555}
}