Intro to computer vision

For those of you that haven’t been paying attention to the latest AI developments, it seems that it’s still up in the air as to exactly when businesses will start to integrate this cutting-edge technology into our day-to-day lives. But, in a world where AI is becoming increasingly commonplace, the first step looks set to be computer vision, AKA image recognition.

What we mean when we talk about ‘computer vision’ is the technology that gives machines or computers the capability to ‘see’ and visually sense the world as humans do. Companies are becoming increasingly interested in utilising image recognition and machine learning to help them on tasks such as auditing, transactional activity, security and marketing – especially when it’s connected to content discovery, hyper-segmentation and customer engagement.

Here are some examples.

Within our grasp – Two significant steps

Now back to reality; how close are we to seeing this technology being introduced into our everyday lives?

The significant barriers to this new technology are the lack of AI capabilities; the algorithms that make it possible to decipher pixels into actionable data, image recognition and a solid use-case that provides the ROI needed to produce a final product.

But that doesn’t mean you can’t tap into this powerful technology right now. Both Google and Amazon have now launched their APIs on image recognition, facilitating the use of this technology by eradicating the need to build your own deep-learning engine. These are relatively complex engines – able to detect a scene, facial recognition and analysis, person tracking, unsafe content detection, celebrity recognition and text-in-images recognition.

There’s no doubt that when it comes to the use of computer vision, the possibilities are endless.

What does this mean for you?

It’s still early days, and the key to any successful digital product is first to uncover a business challenge, then use technology to overcome it, not fixating on a technology and then creating a problem to solve it.

Having said that, there might be a business case to introduce image recognition into your digital products. This technology could be used to enhance existing features like stock control, auditing, security and health & safety which could all be improved in one way or another.

We’ll be working with our clients and partners to make sure they get the most out of this developing technology now that we can add this important tool to our belt of solutions.

Sources:

https://techcrunch.com/2017/11/21/google-lens-in-the-assistant-starts-rolling-out-to-all-pixel-users/

http://www.bmva.org/visionoverview

https://venturebeat.com/2017/11/21/amazon-web-services-computer-vision-gets-real-time-face-recognition/

https://cloud.google.com/vision/

https://aws.amazon.com/rekognition/