Computer Vision Set to Transform Video Contextual Advertising |   10 Feb, 2020

Over the years, contextual video advertising has taken the center-stage in the online advertising space. Video advertising has become the preferred advertising method for many major brands and publishers since it does not rely on third-party cookies.

In fact, 2017 was considered to be the “Video Gold Rush” year for video contextual advertising. This was because when video advertisements were placed in the appropriate setting (context and location), they improved KPIs and enhanced consumer engagement and ROI. 

Videos are the easiest way to grab the attention of your audience. According to stats, the audience reach of online videos is 92%. With this popularity rate, it was also seen that 82% of traffic was derived from videos.

Contextual Video Advertising Makes Strides

It has become essential for marketers to work in secure settings as the quantity of video-based creatives has grown.  There is a greater chance of advertising getting misplaced as reach grows. Misplaced advertisements carry the risk of failing to connect with people, but if they are shown next to undesirable, violent, or divisive material, they may irreparably harm a brand’s reputation.

When in-video advertising proliferated, this resulted in the development of machine learning- and AI-based context detection technologies. These technologies aid in serving ads on the relevant types of content by sorting through all appropriate information that qualifies and categorizing it according to suitability based on geography, brand, and the customers’ state of mind.

The algorithms operate by identifying patterns in the pertinent categories of data later categorizing more recent data in accordance with those patterns. This method has historically been used to categorize ads based on keywords and affinity. 

Natural Language Processing (NLP) techniques have been employed to assess the acceptability of videos for ad placement. Although this method was partly successful in classifying videos into a larger group, it is not impenetrable and has resulted in several leaks in the past. 

The ability of such algorithms to filter content is constrained since the content in videos may not always be relevant to the keywords, search phrases, or descriptions that are linked to them.

Computer Vision is the Key

With the evolution of OTT, Computer vision techniques and their combination with AI and machine learning, a powerful tool is now available to advertisers. Using computer vision and AI, it is possible to identify the actual objects, logos, emotions, and context of the video with unprecedented accuracy. These can enable the identification of what is actually being shown in the video directly without relying on associations with keywords and affinity of videos, thus overcoming the limitation of previously used algorithms, and ensuring a top-notch brand-safe environment for brands.

Major companies like Google, Facebook, Amazon, IBM, and Apple are continuing to bet on Computer Vision for the future and have made considerable advances in the area. Facebook could identify each person from a crowd using facial recognition, while Google can search and index thousands of images through its image search feature. Due to keen interest in the field and momentous advances in the field, computer vision has become increasingly accurate over time. Facebook could identify people with 83% certainty, even when they are partially or completely blocked from view. But in 2022, Facebook disabled the feature due to privacy concerns. 

With these advances, OTT Computer vision has unlocked a whole new avenue for advertisers. With this, one can identify the brand of clothes someone is wearing, the fabric and color he/she prefers, and in what combination, just from their picture. This information can give deeper insights into consumer preferences. In this case, we examine whether luxury goods are a common part of the consumer’s lifestyle and act accordingly. In videos, identifying key aspects of what is being shown can help advertisers in leveraging associations and consumer preferences for brands.

How Brands Benefitted from Mirrors?

Mirrors is an AI-powered contextual targeting platform that uses computer vision applications for logo and object detection, along with facial recognition and emotion identification for classifying videos for placing the most relevant ads in the right context. Working with brands such as Oreo and KFC in the past, we have helped brands target consumers throughout the world to drive sales.

For Oreo, we helped in using the brand’s association with Spider-Man, to target consumers in South-East Asia and placing ads for Oreo in a wide range of Spider-Man-themed content from animated videos to videos showing DIY craft projects for kids. All of this was done in a completely brand-safe environment, with ads being served on videos that connected with the nature of the brand.