{"success":true,"data":{"author":"Konstantin Egorov, Stepan Botman, Pavel Blinov et al.","bucketAccessType":"public","bucketName":"kyegorov/mcd_rppg","bucketPrefix":"","category":"Physiology","cloudProvider":"huggingface","createdAt":"2026-11-17T19:29:32.840Z","description":"MCD-rPPG (Multi-Camera Dataset for Remote Photoplethysmography) is a dataset for estimating pulse waves and health biomarkers from facial video. 600 adult subjects were each recorded twice, at rest and after exercise (15 squats in 30 seconds), for three minutes at a time by three consumer cameras at different angles (frontal webcam, FullHD camcorder and mobile phone), giving 3,600 synchronized videos. Each recording is paired with contact PPG at 100 Hz and ECG, and each subject with 13 biomarkers: systolic and diastolic blood pressure, oxygen saturation, temperature, glucose, glycated hemoglobin, cholesterol, respiratory rate, arterial stiffness, stress level (PSM-25), age, sex and BMI. The repository has video/ (3,600 files), ppg/ and ecg/ (1,200 each), ppg_sync/ and ppg_sync_npy/ (PPG aligned to each video) and meta/, indexed by db.csv; about 130 GB (measured 2026-11-17). Intended for training and benchmarking rPPG models, cross-dataset generalization and multi-task biomarker regression; the authors' baseline code is at github.com/ksyegorov/mcd_rppg. The videos show identifiable faces; the card asks users not to use them for identification or biometric surveillance. The dataset card states the license is CC BY 4.0; cite Egorov et al., Gaze into the Heart, ACM Multimedia 2025 (doi 10.1145/3746027.3758255).","displayTitle":"MCD-rPPG: multi-camera facial video with synchronized PPG, ECG and 13 health biomarkers from 600 subjects","doi":"10.1145/3746027.3758255","huggingFaceUrl":"https://huggingface.co/datasets/kyegorov/mcd_rppg","id":"5599a882-cae8-504b-96d2-ee9ae2978279","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/kyegorov/mcd_rppg","objectCountEstimate":16804,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2508.17924","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"kyegorov/mcd_rppg\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"kyegorov/mcd_rppg\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=kyegorov, license:cc-by-4.0, size_categories:1K<n<10K, downloads=29323"],"format":"mixed","generatedAt":"2026-11-17T19:40:04.484Z","layout":{"example":"db.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"patient_id,weight,height,bmi,age,sex,upper_ap,lower_ap,saturation,temperature,hemoglobin,glycated_hemoglobin,cholesterol,respiratory,rigidity,pulse,stress,step,camera,view,ecg,ppg,video,meta,ppg_sync\n1020,55.0,170.0,19.031141868512112,23.0,F,113.0,78.0,98.0,36.6,12.6,5.56,4.3,19.0,10.79,100.0,4.0,after,FullHDwebcam,front,ecg/1020_after.json,ppg/1020_after.PW,video/1020_FullHDwebcam_after.avi,meta/1020_FullHDwebcam_after.txt,ppg_sync/1020_FullHDwebcam_after.txt\n1020,55.0,170.0,19.031141868512112,23.0,F,113.0","encoding":"utf8","key":"db.csv","truncatedAt":512}],"sampleObjects":["db.csv","ecg/1020_after.json","ecg/1020_before.json","ecg/1024_after.json","ecg/1024_before.json"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntags:\n- rPPG,\n- video,\npretty_name: MCD-rPPG\nsize_categories:\n- 1K<n<10K\n---\n# Dataset Card for Dataset Name\n\nMCD-rPPG (Multi-Camera Dataset for Remote Photoplethysmography) is a large-scale multimodal dataset designed for remote photoplethysmography (rPPG) and health biomarker estimation from facial video.\n## Dataset Details\n\n### Dataset Description\n\nMCD-rPPG (Multi-Camera Dataset for Remote Photoplethysmography) is a large-scale multimodal dataset designed for remote photoplethysmography (rPPG) and health biomarker estimation from facial video. The dataset addresses critical limitations of existing rPPG datasets, including small size, privacy concerns with facial videos, and lack of diversity in recording conditions.\n\nThe dataset comprises 3,600 synchronized video recordings from 600 subjects, captured under varied conditions (resting and post-exercise) using multiple consumer-grade cameras at different angles. Each recording is paired with 100 Hz PPG and ECG signals, along with 13 extended health biomarkers.\n\n- **License:** cc-by-4.0\n\n### Dataset Sources\n\nRepository: https://github.com/ksyegorov/mcd_rppg\n\nHugging Face Hub: https://huggingface.co/datasets/kyegorov/mcd_rppg\n\nPaper: Egorov, K., Botman, S., Blinov, P., Zubkova, G., Ivaschenko, A., Kolsanov, A., & Savchenko, A. (2025). Gaze into the Heart: A Multi-View Video Dataset for rPPG and Health Biomarkers Estimation. Proceedings of the 33rd ACM International Conference on Multimedia. DOI: 10.1145/3746027.3758255\n\narXiv: https://arxiv.org/abs/2508.17924v1\n\n## Uses\n\n<!-- Address questions around how the dataset is intended to be used. -->\n\n### Direct Use\n\nThe dataset is intended for:\n- Training and evaluating rPPG models for pulse wave signal estimation from facial video\n\n- Multi-task learning for simultaneous estimation of PPG signals and health biomarkers\n\n- Cross-dataset evaluation of rPPG model generalization\n\n- Research on contactless health monitoring and telemedicine applications\n\n### Out-of-Scope Use\n\nClinical diagnosis or medical decision-making without validation\n\n- Identification or re-identification of subjects from facial video data\n\n- Commercial deployment without proper validation and ethical review\n\n- Use of facial videos for biometric recognition or surveillance purposes\n\n## Dataset Structure\n\nThe dataset is organized as follows:\n\n    3600 video recordings (600 subjects × 2 states × 3 cameras)\n\n    Synchronized PPG (100 Hz) and ECG signals\n\n    13 health biomarkers: systolic/diastolic pressure, oxygen saturation, temperature, glucose, glycated hemoglobin, cholesterol, respiratory rate, arterial stiffness, stress level (PSM-25), age, sex, BMI\n\n    Multi-view videos: frontal webcam, FullHD camcorder, mobile phone camera\n\nEach subject was recorded twice: before and after physical activity (e.g., 15 squats in 30 seconds). Each measurement corresponds to a three-minute video simultaneously captured from three different angles.\n\n## Dataset Creation\n\n### Curation Rationale\n\nThe dataset was created to address the critical limitations of existing publicly available rPPG datasets:\n\n- Small size: Existing datasets typically contain only 10–140 subjects\n\n- Privacy concerns: Most known datasets require restricted access with no guarantee of approval, impeding collaborative research\n\n- Lack of diversity: Insufficient variation in recording conditions, subjects, and physiological states\n\n- Limited health metrics: Most datasets lack comprehensive health data such as temperature, ECG, and other biomarkers\n\nThe public release of the dataset aims to significantly accelerate progress in the development of AI medical assistants.\n\n### Source Data\n\n\n#### Data Collection and Processing\n\nData was collected using a multimodal capture setup involving three consumer-grade cameras at different angles: a frontal webcam, a FullHD camcorder, and a mobile phone camera. Each subject was recorded in two states: resting and post-exercise (e.g., 15 squats in 30 seconds).\n\nSynchronized PPG signals were sampled at 100 Hz, alongside ECG signals. Extended health m","encoding":"utf8","key":"README.md","name":"README.md","size":7476,"truncated":true},{"bytes":"This is repository for  dataset\n","encoding":"utf8","key":"readme.txt","name":"readme.txt","size":32,"truncated":false}],"sourceObjectKey":"db.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/kyegorov/mcd_rppg/db.csv\")\ndf.head()","starterSamplePath":"db.csv","totalObjects":16804},"summary":"MCD-rPPG (Multi-Camera Dataset for Remote Photoplethysmography) is a dataset for estimating pulse waves and health biomarkers from facial video.","title":"MCD-rPPG: multi-camera facial video with synchronized PPG, ECG and 13 health biomarkers from 600 subjects","updatedAt":"2026-11-17T23:20:55.179Z","visibility":"public","websiteUrl":"https://github.com/ksyegorov/mcd_rppg"}}