MTVital - Efficient Deep Learning-based Estimation of the Vital Signs on Smartphones

Official repository of "Efficient Deep Learning-based Estimation of the Vital Signs on Smartphones".

This repository contains the code and the proposed dataset (MTHS)

Checklist

MTHS dataset
Finger Videos (New!)
Code
Webpage

Code

Codes are included in the code folder. Please refer to its Readme.md for more detailed information.

Dataset - MTHS:

This folder contains our dataset
Each subject has two .npy files: mean RGB signals as signal_x.npy and ground truth labels as label_x.npy, where x is the patient id.
signal_x.npy contains the mean signals ordered as R, G, and then B sampled at 30Hz.
label_x.npy contains the ground truth data ordered as HR(bpm) - SpO2(%) Sampled at 1Hz.

New! - Fingertip videos

Due to some requests we now provide the raw fingertip videos.

For downloading videos, please send us an email with your academic email containing your Gmail address.

Donation :)

If you find our dataset useful please consider donation, it would help us a lot.

btc:

bc1qtgtflqv0laapltmwczfg8ree70mv90fcwvvsd4

eth:

0xCa432902f1270AD076814cD77E03Aef2D09dAc19

usdt (trc20):

TYRPhPT5BZTvn4bcYY4xguL1FSVwchHHrN

License

This project's code is released under the MIT license. Note that the dataset is released under the CC BY-NC-ND license.

Citation

If you use our dataset or find this repository helpful, please consider citing:

@misc{2204.08989,
Author = {Taha Samavati and Mahdi Farvardin},
Title = {Efficient Deep Learning-based Estimation of the Vital Signs on Smartphones},
Year = {2022},
Eprint = {arXiv:2204.08989},
}

gpg-dev/MTVital