The objective of this study, undertaken by researcher Matthias S. Keil from the Basic Psychology Department of the UB and published in the prestigious US journal PLoS Computational Biology, was to ascertain which specific features the brain focuses on to identify a face. It has been known for years that the brain primarily uses low spatial frequencies to recognise faces. "Spatial frequencies" are, in a manner of speaking, the elements that make up any given image.
As Keil confirmed to SINC, "low frequencies pertain to low resolution, that is, small changes of intensity in an image. In contrast, high frequencies represent the details in an image. If we move away from an image, we perceive increasingly less details, that is, the high spatial frequency components, while low frequencies remain visible and are the last to disappear."
As a result of the psychophysical research carried out prior to the publication of this study, it was known that the human brain was not interested in very high frequencies when identifying faces, despite such frequencies playing a significant role in, for example, determining a person's age. "In order to identify a face in an image, the brain always processes information with the same low resolution, of about 30 by 30 pixels from ear to ear, ignoring distance and the original resolution of the image," Keil says. "Until now, nobody had been able to explain this peculiar phenomenon and that was my starting point".
What Matthias S. Keil did was to analyse a large number of faces, namely those belonging to 868 women and 868 men. "The idea was to find common statistical regularities in the images." Keil used a model of the brain's visual system, that is, "I looked at the images to certain extent like the brain does, but with one difference: I had no preferred resolution, but considered all spatial frequencies as equal. As a result of this analysis, I obtained a resolution that is optimum in terms of encoding, as well as the signal-to-noise ratio, and was also the same resolution observed in the psychophysical experiments".
This result therefore suggests that faces are themselves responsible for our resolution preference. This led Keil to one of the brain's properties: "The brain has adapted optimally to draw the most useful information from faces in order to identify them. My model also predicts this resolution if we take into account the eyes alone – ignoring the nose and the mouth – but also by considering the mouth or nose separately, albeit less reliable."
Therefore, the brain extracts key information for facial identification primarily from the eyes, while the mouth and the nose are secondary, according to the study. According to Keil, if we take a photo of a friend as an example, one might think that every feature of the face is important to identify the person. However, numerous experiments have demonstrated that the brain prefers a coarse resolution, regardless of the distance between the face and the beholder. Until now, the reason for this was unclear. The analysis of the pictures of 868 men and 868 women in this study could help to explain this.
The results obtained by Kiel indicate that the most useful information is drawn from the images if they are around 30 by 30 pixels in size. "Furthermore, the pictures of the eyes provide the least 'noisiest' result, which means that they transmit more reliable information to the brain than the pictures of the mouth and the nose," the researcher said. This suggests that the brain's facial identification mechanisms are specialised in eyes.
This research complements a previous study published by Keil in PLoS ONE, which already advanced that artificial face identification systems obtain better results when they process small pictures of faces, which means that they could behave in this sense like humans.
References: Mathias S. Keil. "I Look in Your Eyes, Honey: Internal Face Features Induce Spatial Frequency Preference for Human Face Processing". PLoS Computational Biology. número 5(3), marzo de 2009.
SINC | EurekAlert!
Drone vs. truck deliveries: Which create less carbon pollution?
31.05.2017 | University of Washington
New study: How does Europe become a leading player for software and IT services?
03.04.2017 | Fraunhofer-Institut für System- und Innovationsforschung (ISI)
An international team of scientists has proposed a new multi-disciplinary approach in which an array of new technologies will allow us to map biodiversity and the risks that wildlife is facing at the scale of whole landscapes. The findings are published in Nature Ecology and Evolution. This international research is led by the Kunming Institute of Zoology from China, University of East Anglia, University of Leicester and the Leibniz Institute for Zoo and Wildlife Research.
Using a combination of satellite and ground data, the team proposes that it is now possible to map biodiversity with an accuracy that has not been previously...
Heatwaves in the Arctic, longer periods of vegetation in Europe, severe floods in West Africa – starting in 2021, scientists want to explore the emissions of the greenhouse gas methane with the German-French satellite MERLIN. This is made possible by a new robust laser system of the Fraunhofer Institute for Laser Technology ILT in Aachen, which achieves unprecedented measurement accuracy.
Methane is primarily the result of the decomposition of organic matter. The gas has a 25 times greater warming potential than carbon dioxide, but is not as...
Hydrogen is regarded as the energy source of the future: It is produced with solar power and can be used to generate heat and electricity in fuel cells. Empa researchers have now succeeded in decoding the movement of hydrogen ions in crystals – a key step towards more efficient energy conversion in the hydrogen industry of tomorrow.
As charge carriers, electrons and ions play the leading role in electrochemical energy storage devices and converters such as batteries and fuel cells. Proton...
Scientists from the Excellence Cluster Universe at the Ludwig-Maximilians-Universität Munich have establised "Cosmowebportal", a unique data centre for cosmological simulations located at the Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Sciences. The complete results of a series of large hydrodynamical cosmological simulations are available, with data volumes typically exceeding several hundred terabytes. Scientists worldwide can interactively explore these complex simulations via a web interface and directly access the results.
With current telescopes, scientists can observe our Universe’s galaxies and galaxy clusters and their distribution along an invisible cosmic web. From the...
Temperature measurements possible even on the smallest scale / Molecular ruby for use in material sciences, biology, and medicine
Chemists at Johannes Gutenberg University Mainz (JGU) in cooperation with researchers of the German Federal Institute for Materials Research and Testing (BAM)...
19.06.2017 | Event News
13.06.2017 | Event News
13.06.2017 | Event News
26.06.2017 | Life Sciences
26.06.2017 | Physics and Astronomy
26.06.2017 | Information Technology