Proprietary Computer Vision 2.0

Computer Vision is one of the most important technologies to power digital transformation. Using computer vision many tasks can be performed in extremely efficient and scalable manner:

Using computer vision in telecom sector, for example, a given company can easily build mobile apps that use phone cameras to detect configuration of equipment, do automatically inventory, compare states of the equipment, detect anomalies. The same algorithms can be used to detect equipment and overlay augmented reality content or data on top of the equipment.

So why is it not done yet? Because you need proprietary data for computer vision to understand what is on the picture. And it is not being collected due to high effort and cost..

Collecting those images is an important task strategically because it gives a competitive advantage, however, it is comes with large cost for industries with low cash on their hands. And you need images almost per each possible situation, light, angle, and etc. A standard way is to create a program or a project, align resources and create a task to collect images. Running such a large scale task in traditional way creates a lot of costs.

The trick here is that Bitskout makes the task non-obligatory and add extra motivation to snap a photo while you are at a location doing other things. This direct execution - reward relationship combined with an automated validation, payment and storage of the submitted data creates a perfect combination to boost motivation and collect proprietary data. Here is a video of how it works:

Above example is a way to use campaigns to distribute volume based tasks to your employees avoiding IP issues in case of using external crowdsourcing and boosting their motivation by allowing them directly be part of new things.

It means that not only you get an important proprietary data asset, you also motivate your people. And motivated employees can do wonders :

Cheers,

Ilia Zelenkin