[attach id=598937 size="medium" align="right"]These images show an average of 500 images of male faces (left) and 500 images of female faces (right). Computer systems can use these averages as a means of detecting the features that can be used to classify a person’s gender.[/attach]
From a visual perception point of view, the face is one of the most interesting parts of the human body. It has some distinctive physical and expressive features that allow us to draw some conclusions about an individual. For instance, by just looking at the face of someone, we can recognise the gender and the ethnicity, make a rough estimation of the age, analyse the expression so as to deduce the main emotion and determine if the person has a familiar face or is a stranger.
Although all faces consist of the same components in a specific spatial arrangement (the relative positions of the nose, eyes and others), primates have enviable abilities to use the subtle features and infer certain states from faces in a remarkable seemingly and effortless operation.
As a matter of fact, there is neurophysiological evidence that the visual cortices of primates have single neurons that respond to faces. This fact demonstrates the importance faces were given by evolution. The images illustrate the average faces of 500 adult men and 500 adult women of different ethnicity. One may observe the subtle differences between the two genders. For instance, women seem to have rounder and lighter skin faces, bigger cheekbones, as well as narrower noses and eyebrows.
In recent years, the variety and appeal of faces motivated several research communities to work on the automatic face analysis from images and videos. In particular, gender recognition has been considered among the most popular and challenging face analysis problem.
The computer programmes for the automatic classification of gender have a lot of potential for commercial applications. Imagine being in a taxi in London or New York, where most of them are equipped with passenger screens. A small camera could be used to take snapshots of your face followed by automatic face analysis.
The determined properties, such as gender, age, ethnicity and emotion, could then be used to personalise any advertisements. This concept could be used in other transport means, such as trains and planes. Other applications of face analysis could be used in shopping malls to collect statistics about the customers, and to improve the content-based indexing of faces, among others.
Research at the University of Malta, in collaboration with the MIVIA Lab of the University of Salerno (Italy), is looking at the development of an embedded system that performs face analysis. It is investigating novel computer vision techniques that can extract the subtle visual features and use them to build models that can distinguish the gender, age group, and ethnicity. The research has produced a novel method, which was published in the international AVSS conference, and demonstrated that their approach outperforms commercial software.
Dr George Azzopardi is a lecturer with the Department of Artificial Intelligence at the University of Malta.
Did you know?
• The typical brain is about two per cent of a body’s weight but uses 20 per cent of its total energy and oxygen intake.
• Your brain is 73 per cent water. It takes only two per cent dehydration to affect your attention, memory and other cognitive skills.
• Twenty-five per cent of the body’s cholesterol resides within the brain. Cholesterol is an integral part of every brain cell. Without adequate cholesterol, brain cells die.
• A piece of brain tissue the size of a grain of sand contains 100,000 neurons and one billion synapses all communicating with each other.
• As any parent can attest, teenage brains are not fully formed. It isn’t until about the age of 25 that the human brain reaches full maturity.
• The average brain is believed to generate around 50,000 thoughts per day. Disturbingly, it’s estimated that in most people 70 per cent of these thoughts are negative.
For more trivia see: www.um.edu.mt/think
Sound bites
• Physical exercise has an anti-ageing effect on the hippocampus region of the brain – an area that controls memory, learning and balance. A new study, comparing different forms of exercise – dancing and endurance training – undertaken by elderly volunteers for 18 months shows that both can have an anti-ageing effect on the brain, but only dancing corresponded to a noticeable difference in behaviour. This difference is attributed to the extra challenge of learning dancing routines.
https://www.sciencedaily.com/releases/2017/08/170825124902.htm
• Our gut instinct about whether a stranger poses a threat is as good when we’re 80 as when we’re 18, according to new research. Older people are as good as young adults at knowing when someone is potentially aggressive, and being streetwise appears to be a skill honed in childhood but not fully reliable until adulthood.
https://www.sciencedaily.com/releases/2017/08/170828093743.htm
For more soundbites listen to Radio Mocha on Radju Malta 2 every Monday at 1pm and Friday at 6pm https://www.facebook.com/RadioMochaMalta/