Performance Evaluation of rPPG Approaches with and without the Region-of-Interest Localization Step

Pirnar, Žan; Finžgar, Miha; Podržaj, Primož

doi:10.3390/app11083467

Cited by 4 publications

(3 citation statements)

References 46 publications

(104 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Skin segmentation: Moreover, skin segmentation can be applied to catch skin pixels only without any background pixels [67,68]. • 1-D signal conversion: The video sequence is converted into a 1-D signal by averaging the intensity levels of the whole frame pixels or the localized pixels in the intended ROI [69]. So, we have three independent vectors arising from RGB channels.…”

Section: Camera-based Rppg Extraction Overviewmentioning

confidence: 99%

Video-based beat-by-beat blood pressure monitoring via transfer deep-learning

Omer,

Salah,

Hassan

et al. 2024

Appl Intell

View full text Add to dashboard Cite

Currently, learning physiological vital signs such as blood pressure (BP), hemoglobin levels, and oxygen saturation, from Photoplethysmography (PPG) signal, is receiving more attention. Despite successive progress that has been made so far, continuously revealing new aspects characterizes that field as a rich research topic. It includes a diverse number of critical points represented in signal denoising, data cleaning, employed features, feature format, feature selection, feature domain, model structure, problem formulation (regression or classification), and model combinations. It is worth noting that extensive research efforts are devoted to utilizing different variants of machine learning and deep learning models while transfer learning is not fully explored yet. So, in this paper, we are introducing a per-beat rPPG-to-BP mapping scheme based on transfer learning. An interesting representation of a 1-D PPG signal as a 2-D image is proposed for enabling powerful off-the-shelf image-based models through transfer learning. It resolves limitations about training data size due to strict data cleaning. Also, it enhances model generalization by exploiting underlying excellent feature extraction. Moreover, non-uniform data distribution (data skewness) is partially resolved by introducing logarithmic transformation. Furthermore, double cleaning is applied for training contact PPG data and testing rPPG beats as well. The quality of the segmented beats is tested by checking some of the related quality metrics. Hence, the prediction reliability is enhanced by excluding deformed beats. Varying rPPG quality is relaxed by selecting beats during intervals of the highest signal strength. Based on the experimental results, the proposed system outperforms the state-of-the-art systems in the sense of mean absolute error (MAE) and standard deviation (STD). STD for the test data is decreased to 5.4782 and 3.8539 for SBP and DBP, respectively. Also, MAE decreased to 2.3453 and 1.6854 for SBP and DBP, respectively. Moreover, the results for BP estimation from real video reveal that the STD reaches 8.027882 and 6.013052 for SBP and DBP, respectively. Also, MAE for the estimated BP from real videos reaches 7.052803 and 5.616028 for SBP and DBP, respectively. Graphical abstract Proposed camera-based blood pressure monitoring system

show abstract

Section: Camera-based Rppg Extraction Overviewmentioning

confidence: 99%

Video-based beat-by-beat blood pressure monitoring via transfer deep-learning

Omer,

Salah,

Hassan

et al. 2024

Appl Intell

View full text Add to dashboard Cite

show abstract

“…The evaluation metrics used to assess the performance of the proposed robot-aided contactless monitoring system consist of the following indicators: in [42], the absolute error is a performance accuracy indicator of the PPG methods and it is calculated as follows:…”

Section: Evaluation Metricsmentioning

confidence: 99%

Robot-Aided Contactless Monitoring of Workers’ Cardiac Activity in Hazardous Environment

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Selecting appropriate ROI can effectively enhance the ruggedness of heart rate detection by IPPG [19][20]. The ROI ought to be capillary-rich, for instance, the frontal region of the face, the cheeks on the left and right, the lips and the chin, etc.…”

Section: Regions Of Interest Acquisitionmentioning

confidence: 99%

A Non-contact Heart Rate Measurement Approach Based on Modified EEMD Algorithm with Updated White-noise Iterating

Shi

Qiu

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

At present, Imaging Photoplethysmography (IPPG) has attracted extensive attention on heart rate measurement, because of its non-contact, non-invasive and safe characteristics. However, in reality, the IPPG signal is extremely feeble and susceptible to be submerged by ambient interference such as illumination. To this end, we propose a new IPPG signal processing approach to enhance the anti-interference capability of measuring heart rate through a modified EEMD algorithm with updated white-noise iterating. In the modified EEMD algorithm, based on the standard deviation of the input IPPG signal, zero-mean white noises are generated randomly for iterative ensemble averaging. Through this new method, the fault caused by the residual white noise of traditional EEMD is solved. Lastly, through the comparative experiments of our own acquisition statistics, the approach proposed by this paper has significantly improved the performance in terms of accuracy, timeliness and anti-interference under various scenarios. The mean absolute error reaches 1.4 beats per minute and the correlation coefficient r reaches 0.8089. In addition, the modified EEMD algorithm reduces the running time by more than 10 times compared to the original method, approximately 0.16s per time.

show abstract

Performance Evaluation of rPPG Approaches with and without the Region-of-Interest Localization Step

Cited by 4 publications

References 46 publications

Video-based beat-by-beat blood pressure monitoring via transfer deep-learning

Video-based beat-by-beat blood pressure monitoring via transfer deep-learning

Robot-Aided Contactless Monitoring of Workers’ Cardiac Activity in Hazardous Environment

A Non-contact Heart Rate Measurement Approach Based on Modified EEMD Algorithm with Updated White-noise Iterating

Contact Info

Product

Resources

About