Synthetic Generation of Face Videos with Plethysmograph Physiology

Wang, Zhen; Ba, Yunhao; Chari, Pradyumna; Bozkurt, Oyku Deniz; Brown, Gianna; Patwa, Parth; Vaddi, Niranjan; Jalilian, Laleh; Kadambi, Achuta

doi:10.1109/cvpr52688.2022.01993

Cited by 28 publications

(15 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, video files from UCLA dataset [13] were used to extract physiological in-formation from individuals. Each video is processed using the proposed method described in section 2.1, which returns the rPPG signal.…”

Section: Resultsmentioning

confidence: 99%

“…For instance, Dasari et al [12] proposed a dataset that only contains dark skin tones, but the actual videos are not shared, only the color space values of the skin region of interest. Despite these challenges, Wang et al [13] recently proposed the largest known rPPG dataset, which includes a variety of participants with different skin tones. The dataset comprises 98 subjects and 489 videos of various skin tones, ages, genders, ethnicities, and races.…”

Section: Ucla Datasetmentioning

confidence: 99%

See 1 more Smart Citation

Evaluation of Remote Monitoring Technology across different skin tone participants

Talukdar

Deus

Sehgal

2023

Preprint

View full text Add to dashboard Cite

Background: Many research studies seek to improve vital sign monitoring to enhance the conditions under which doctors and caregivers track patients health. Non-invasive and contactless monitoring has emerged as an optimal solution for this problem, with telemedicine, self-monitoring, and well- being tools being the next generation of technology in the biomedical field. However, there is worldwide concern about the general purpose and bias towards a certain demographic group of these techniques. In particular, skin tone and the accuracy of monitoring dark skin tone groups have been key questions among researchers, with the lack of results and studies contributing to this uncertainty. Methods: This paper proposes a benchmark for remote monitoring solution against a medical device across different skin tone people. Around 330 videos from 90 different patients were analysed, and Heart Rate and Heart Rate Variability were compared across different subgroups. The Fitzpatrick scale (1-6) was used to classify participants into three skin tone groups: 1 and 2; 3 and 4; 5 and 6. Results: The results showed that our proposed methodology was able to estimate heart rate with a mean absolute error of 3 bpm across all samples and subgroups. Moreover, for Heart Rate Variability (HRV) metrics, we achieved the following results, in terms of Mobility Assistive Equipment (MAE): HRV-IBI (Inter-Beat-Interval) of 10 ms; HRV-SDNN (Standard Deviation of Normal to Normal heartbeats) of 14 ms and HRV-RMSSD (Root Mean Square of Successive Differences between normal heartbeats) of 22 ms. No significant performance decrease was found for any skin tone group, and there was no error trend towards a certain group. Conclusions: The study showed that our methodology meets acceptable agreement levels for the proposed metrics and is well-suited for users who want to understand their general health and wellness. Furthermore, the experiments showed that skin tone had no impact on the results, which remained within the same range across all groups.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Ucla Datasetmentioning

confidence: 99%

Evaluation of Remote Monitoring Technology across different skin tone participants

Talukdar

Deus

Sehgal

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The input PPG signal for the avatars is generated through the convolution of a Gaussian window with the beat sequence, derived from a heart rate frequency range. In contrast to the intricate pipeline of SCAMPS, Wang et al (2022) presents a more user-friendly approach for generating synthetic videos. This method employs a statistical 3D head model, extracting facial features from publicly available in-the-wild face datasets [BUPT-Balancedface ( Wang et al, 2019 )], while the PPG waveforms are recorded from real human subjects.…”

Section: Related Workmentioning

confidence: 99%

Simulating cardiac signals on 3D human models for photoplethysmography development

Wang,

Chahl

2024

Front. Robot. AI

View full text Add to dashboard Cite

Introduction: Image-based heart rate estimation technology offers a contactless approach to healthcare monitoring that could improve the lives of millions of people. In order to comprehensively test or optimize image-based heart rate extraction methods, the dataset should contain a large number of factors such as body motion, lighting conditions, and physiological states. However, collecting high-quality datasets with complete parameters is a huge challenge.Methods: In this paper, we introduce a bionic human model based on a three-dimensional (3D) representation of the human body. By integrating synthetic cardiac signal and body involuntary motion into the 3D model, five well-known traditional and four deep learning iPPG (imaging photoplethysmography) extraction methods are used to test the rendered videos.Results: To compare with different situations in the real world, four common scenarios (stillness, expression/talking, light source changes, and physical activity) are created on each 3D human. The 3D human can be built with any appearance and different skin tones. A high degree of agreement is achieved between the signals extracted from videos with the synthetic human and videos with a real human-the performance advantages and disadvantages of the selected iPPG methods are consistent for both real and 3D humans.Discussion: This technology has the capability to generate synthetic humans within various scenarios, utilizing precisely controlled parameters and disturbances. Furthermore, it holds considerable potential for testing and optimizing image-based vital signs methods in challenging situations where real people with reliable ground truth measurements are difficult to obtain, such as in drone rescue.

show abstract

“…Synthetic data present the possibility to generate data without the expensive costs, and it has the advantage of controlling the properties of the dataset, which can help create the less biased models. [ 110 ] Later, Wang et al [ 111 ] introduced a scalable physics‐based learning model to generate synthetic rPPG videos with diverse attributes such as skin color, and lighting conditions to improve the performance.…”

Section: Future Prospectsmentioning

confidence: 99%

Recent Progress of Optical Imaging Approaches for Noncontact Physiological Signal Measurement: A Review

Zhang

et al. 2023

Advanced Intelligent Systems

View full text Add to dashboard Cite

In recent years, optical imaging techniques have gained wide recognition for the measurement of vital signals, such as heart rate, respiratory rate, oxygen saturation, and blood pressure, which are crucial indicators for evaluating human health conditions in clinical examinations. There is a wide range of optical imaging methods for remote physiological signal monitoring, including RGB imaging, thermal imaging, hyperspectral imaging, depth imaging, and multimodal imaging, which provide spatial information compared to other noncontact measurement approaches, thereby allowing extensive applications in this area. In this survey, some fundamental knowledge about optical imaging methods for vital signal measurement is reviewed, including principles of various optical imaging techniques, processing methods for data analysis, discussion on advantages and disadvantages, application summary, and future prospects. This is a comprehensive overview of the noncontact physiological signal measurement of optical imaging approaches.

show abstract

Synthetic Generation of Face Videos with Plethysmograph Physiology

Cited by 28 publications

References 28 publications

Evaluation of Remote Monitoring Technology across different skin tone participants

Evaluation of Remote Monitoring Technology across different skin tone participants

Simulating cardiac signals on 3D human models for photoplethysmography development

Recent Progress of Optical Imaging Approaches for Noncontact Physiological Signal Measurement: A Review

Contact Info

Product

Resources

About