The Coronavirus ripple is felt across the global economy and the media & advertising sector is no different. As brands and platforms rapidly add terms associated with the disease to their keyword block, Coronavirus has become one of the most blocked keywords. The pace of the blocking is poised to accelerate, as more advertisers adopt the block in their lists and the volume of content about the outbreak grows.
Video content is facing similar issues. Though YouTube announced in February that it would demonetize Coronavirus related videos, then reversed the decision last week. Advertisers looking to avoid unsafe brand exposure around this sensitive topic, are forced to exclude news entirely from their list of targeted channels and publishers.
However, with the current scenario of social distancing and self-imposed isolation, users are voraciously consuming news-related content. Excluding news and related channels entirely from advertising strategies across platforms is not only impacting publisher ad revenues but also killing reach and monetization.
And as the ad inventory on these channels becomes under-priced, advertisers looking to benefit from this would require a lot of work to accurately determine whether a story is about something a brand would like to avoid completely or is it rather something more benign.
How Can Context Relevance Help?
One of the key challenges with keyword blocklists is the inability to identify the right context, especially for video content. In-video context, on the other hand, offers a high degree of context relevance that surpasses limitations of traditional keyword-based blocking. The ability of a context detection system, powered through computer vision, to identify what the video content actually features rather than how it is described can help advertisers set harmful stories apart from benign ones.
What is Computer Vision?
Computer vision is an AI-powered technology that imbibes the complexity of the human vision system, enabling computers to ‘see’ i.e. identify and process objects in visual content including images and streaming videos. Owing to AI advancements, computer vision has even surpassed humans in detecting certain objects.
Most computer vision algorithms can just detect objects, mostly within static images. But more advanced algorithms have been able to accurately identify people, facial expressions, activities, scenes, and even emotions, not only in static images but also in streaming video content.
In-video contexts to avoid unsafe brand exposure related to Coronavirus
- On-screen text recognition: The computer vision system can identify and filter out videos that have related text written on the screen (e.g. Coronavirus or COVID-19). And can help distinct between serious stories Vs. stories around precautions and more.
- Object and action detection: the system can identify objects like masks, stretchers, and actions like coughing and sneezing and understand the concentration of this content within a single video through frame by frame parsing.
- Faces: with this outbreak certain public figures are also on brands’ blocklists (yes, Trump). In-video context detection can accurately identify faces to filter out related content.
In-video Context Detection
Computer vision’s AI-powered in-video context detection can accurately block ad placements against unwanted, unsuitable, irrelevant and harmful content. These solutions can provide the highest accuracy, not letting damaging ad placements pass through by using frame by frame parsing of video content.
In-video context detection further enables offering true brand suitability. Every brand is different and so are specific brand safety needs. Computer vision offers a tailored approach, offering absolute brand control: unlike keyword blacklists and white-listed channels, that likely also block perfectly safe content.
Dynamic brand safety, with higher and more relevant reach
In-video context detection opens a whole new set of audience to improve reach, with unparalleled brand safety. The ad has a higher probability to match its environment in terms of context and messaging. It runs on the principal that users are engaging with their interests while consuming certain content, and engaging at the right moment can augment this experience and gain interest and trust.