Originally published in News Medical Life Sciences on June 19, 2019.
New technology by Neurodata Lab is capable to acquire physiological data by simply pointing a webcam at one’s face, with accuracy comparable to special wearable equipment.
In the last few years, various research groups developed an alternative to wearable technology capable to measure heart rate remotely, via a video camera. Until very recently, such technology only worked well on HQ video but required significant investments in storage and transmission, which limited their real-life applications.
The technology by Neurodata Lab accurately detects heart rate even in low-quality videos. It extracts a range of physiological characteristics by tracking pixel color change on the head, face, and neck via a webcam in any type of shooting conditions: compressed video stream, variable illumination, presence of body large and small movements, with makeup and different clothing, etc.
In a small in-house study, this remote heart rate detection algorithm was compared to the most popular wearables that can measure pulse, and it turned out to be more accurate than most contact analogs – only 2.38 BPM different as compared to professional wearable equipment.
The core value of the method is in its potential application ‘in the wild’. That is why we test our algorithms in the situations where people move, talk, experience different emotional states.”
Olga Perepelkina, Chief Research Officer at Neurodata Lab