Computer vision is a branch of artificial intelligence (AI) that aims to create techniques that enable computers to see and understand visual content such as images and videos. 

When processing visual content, computers look for patterns and link them to labels that make sense for us humans. For example, if the machine is taught that two eyes, a nose and a mouth, are features of a face, it will tell us that a face was detected in an image when those features or patterns were found. The same goes for any other object.

Consider a scene where the computer programme detects different objects such as a person and a car. During the past decade, we have seen countless methods that successfully find objects in a scene and tell us what they are. For example, computers can successfully detect the person and the car, draw a bounding box around them and even tell us which pixels represent the person and the car.

While computers can successfully detect objects, they do not indicate the order in which we can interpret the scene. However, this advancement introduces a new problem in computer vision: answering questions such as “Which object in the scene attracts the most attention?”

Saliency detection techniques tell us which pixels or regions in an image stand out and draw the most attention. It is easy to decide which object attracts the most attention when only a single object is in a scene. The challenge arises when there is more than one object in a scene, and this introduces the need for a more advanced technique.

Saliency detection techniques tell us which pixels or regions in an image stand out and draw the most attention

Saliency-driven object ranking is an approach designed and built at the University of Malta that allows for ranking objects in a scene based on how much they stand out in the image. The development of this approach was part of the author’s PhD thesis under the supervision of Carl J. Debono, a professor and engineer.

Saliency is used in our ranking method to translate this ranking of importance to objects, allowing the computer to understand which objects in a scene have a higher priority. The potential applications of this new approach to object detection are various. This method can enhance product placement by telling us which product draws the most attention on a shelf. It can also tell us which part of a user interface attracts interaction prior to user trials. It can also allow us to create more meaningful descriptions of the world around us for the blind.

Dylan Seychell is a lecturer in the Department of Artificial Intelligence at the University of Malta.  His PhD studies were supported by the Malta Government Scholarship Scheme (MGSS).

Sound Bites

•        Researchers have made a genetic discovery that could spell the end of disease-carrying mosquitoes and prevent them from maturing or multiplying. Prior to this discovery, the hormone ecdysone was known to act as a transporter for reproduction in insects. However, mosquitoes have only three of the four transporter proteins, missing the ecdysone protein. This discovery opens the door to a mosquito-specific insecticide that would not harm bees and other pollinators.

•        Researchers at Cambridge studying cognition, behaviour and the brain have concluded that people with dyslexia are specialised to explore the unknown and that this ‘explorative bias’ has an evolutionary basis playing a crucial role in our survival. Researchers believe that there is a cognitive trade-off between exploration of new information and exploration of existing information/.

For more sound bites listen to Radio Mocha www.fb.com/RadioMochaMalta/.

DID YOU KNOW?

•        African lions can grow to measure three metres from head to tail, with the tail being about 60 to 90 centimetres long.

•        A lion can typically weigh between 150 to 250 kilos, with males being on the higher end of the range.

•        The bigger the mane of a male lion is, the more intimidating it is to competing males and more attractive to female lions.

•        Females are the main hunters of the pride and work cooperatively in hunting parties to surround and take down prey.

For more trivia, see: www.um.edu.mt/think.