Advanced image recognition tech makes self-drive cars even smarter!

7 Jun

It’s not just encrypted data backups that are the talk of the town these days – in fact, self-drive cars have also been steering some hot debate, particularly in recent months.  Just how safe are they, really?  What are the risks and dangers involved?  These are just two of the many questions that people are asking.

Researchers have been doing some impressive work towards putting the doubts and fears of sceptics to rest – although there are still questions and concerns surrounding the concept of trusting a self-drive vehicle.

Objective and human-like decision-making

The latest in self-drive care development is image-recognition technology.  According to the experts, when incorporated into a self-drive car system this new technology will enable the car to learn to respond to unfamiliar objects on the road, much like a driver; recognising –“this looks like a cyclist” and responding accordingly.

This particular type of image recognition could make all the difference when using a self-drive car, especially in conditions where visibility is poor.  Self-drive car concepts are known to be able to recognise fixed objects as a result of data they draw from deep neural networks.  In the past, anything unexpected or out of the ordinary represented a potential problem or risk.

Jonas Kubilius and Hans Op de Beeck, researchers from the Belgium University KU Leuven working on the self-drive care image recognition technology, say that the more advanced systems will make self-drive cars smarter.  They will be able to recognise objects and obstacles with greater accuracy and also learn human-type sensitivities to object shapes that aren’t clearly visible.


Don’t quite understand the concept?  Here’s what Jonas Kubilius had to say about it: “Machines can learn to tell us what a new shape — say, a letter from a novel alphabet or a blurred object on the road — reminds them of.  This means we are on the right track in developing machines with a visual system and vocabulary as flexible and versatile as ours”.

While it would seem that the self-drive concept has great things in store for it, Kubilius did imply that self-drive cars and their technology cannot, especially in their current phase, be fully trusted to handle all driving responsibilities with his statement: “We are not there just yet.  And even if machines will at some point be equipped with a visual system as powerful as ours, self-driving cars would still make occasional mistakes”.

These developments alone show us just how far technology and connectivity has come in the last few years – months even!  Self-drive cars – mind-controlled drones, what next?!

We’d like to know what you think about self-drive cars and the latest developments in image-recognition technology.  Let us know!

Comments are closed.