Lorem ipsum dolor sit amet, consectetur adicing elit ut ullamcorper. leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet. Leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet.

  /  Project   /  Blog: Object Detection — A Game Changer for Market Research

Blog: Object Detection — A Game Changer for Market Research

“Object Detection” is a branch of Computer Vision that deals with finding specific objects (like humans, RedBull Cans, cartons of RedBull Cans etc.) from an image. With this blog, we will make a case about why Object Detection can be a game changer for market researchers.

Computer Vision is the field that deals with empowering computers the ability to ‘see’ things like humans. Object Detection is a basic visual perception task and one of the key areas of applications of Computer Vision. It essentially deals with finding and locating specific objects within an image.

For detecting generic objects (like a car, person, table, tree) there are open-source and pre-trained models like Yolo available. However, if you want an algorithm to detect very specific objects (like a ‘small raw tomato’ or a ‘large ripe tomato’), you will need to train an object detection algorithm of your own.


Take this video for example. In this experiment, our team was able to perform AI shelf analysis and identify all the red bull cans while a shopper strolls through a retail store. This is just one way in which market research can leverage the power of object detection for better consumer insights.

Another very important application can be in the field of retail audits.

Conducting retail audits is necessary to remain competitive in an industry where multiple products and brands fight for the attention of consumers but even today retail audit is a manual-intensive process. Hundreds of field agents roam around retail stores collecting shelf data and after compilation, they share it with the management. It sounds straightforward and simple, but it is anything but simple. The time-lag between a collection of data from retail stores until a final analysis is substantial which leads to companies losing time to take corrective action. By the time they realize the problem and formulate the solution, it is too late and the companies lose out on sales.

Data accuracy is another concern for management teams. They have to rely solely on the wisdom and honesty of the field agents to collect accurate data.

In today’s environment what companies need is a complete, end‐to‐end retail audit solution that would enable them to improve their shelf presence in real time. It is imperative for companies to fill the time gap between data collection and final analysis. This is where Object Detection comes to the rescue!

Using object detection for retail audits, companies will be able to improve the timeliness and accuracy of reports that will enable their managers to take in time corrective action.


Eye-Tracking is another very interesting field of market research that leverages the power of object detection.

In the video, you can see how visual input from eye tracking glasses has been coded to give insights about consumer behaviour inside a store. Eye tracking empowered with object detection can have a myriad of applications in market research such as Pack Testing, Advertisements, consumer behaviour analysis and many more.

In summary Object detection can help companies get important information about a number of key areas such as Actual requirement of a customer, Customer behaviour, his needs, habits, preferences, Main motivators and barriers at retail points, Optimize the point of sale (POS), Role of displays to the customers and many more.

We at Karna AI are leveraging the power of AI to bring object detection to practical use by integrating it with our market research solutions such as Smart Gaze for Eye-Tracking and Shelf Watch, our Shelf-Monitoring solution.

Want to improve your market research? Click here to schedule a free demo.

Source: Artificial Intelligence on Medium

(Visited 2 times, 1 visits today)
Post a Comment