Identifying specific product using Computer Vision

Through product advertisements (TV, print and digital), companies seek to attract their customers’ attention and influence consumer buying behaviour. FMCG companies produce hundreds of advertisements regularly across their product portfolio, and for each and every advert, high quality advertising industry standards must be adhered to. So being able to run stringent quality checks at speed is imperative. In this experiment, we explored a deep learning solution to examine a given ad against the criteria set and to check whether the ad would pass the stringent quality check mandates for the ads set by the companies. Our goal with this experiment is to help brands automate their entire quality checking process and reduce manual efforts in ad checking.

Background

Computer vision has seen significant growth in this century owing to advances in AI – specifically deep learning and neural network innovations. Through computer vision, AI systems can recognise and make sense of the world using cameras, videos and images, similar to how humans use their eyes and brains to understand the world around them. Computer vision algorithms not only detect objects but can translate visual content into meta-data for further analysis.

Digital systems set up to evaluate ads process them using deep learning and AI and then identify the objects and components within them. One of the tasks at hand is to locate the product in a given ad and track it in the video of the ad when in motion. The hindrances to this are motion blur and variable aspect ratios which are also some of the common issues in working with video formats. Reading the product brand name is challenging as well, since the product under evaluation can be in any position or orientation. To tackle all these challenges, we have made a computer vision-based solution that detects the product precisely and checks its branding as well.

Business Use Cases and Applications

There are multiple applications of this experiment. Some of them include –

1. Branded Object Recognition: Brands have to track customers’ behavior on various social media platforms to identify their engagement with their products. The approach of this experiment helps brands see the threats and opportunities that can influence their success.

2. Shelf Monitoring in Retail Stores: Computer vision is being used to improve the efficiency in retail stores. With the help of smart cameras and sensor systems, similar solutions can be developed to track the number of products available in the shelves in real-time.

3. Self checkout/Amazon Go stores: This product detection solution can also be extended to understand what products the customers are picking from the shelf, thus enabling automated checkout.

To enjoy more of this content, please enter your details to receive full access.

Do you have an AI use case you want to explore?

If you’ve got a specific AI use case that you’d like some help exploring, if you’re interested in collaborating with us as partners, or if you’re just interested in finding viable and effective ways to apply AI in your organisation, speak to us today.

hbspt.forms.create({ region: "na1", portalId: "2495356", formId: "3c5c77b2-406f-4cd8-a872-c3aa17c17f73" });