SilverPush has announced the official launch of Mirrors, the company’s artificial intelligence (AI)-driven context detection technology that identifies brand logos, faces, emotions and objects in videos. By using AI with computer vision, Mirrors detects context in video content that aligns with the advertiser’s core communications objectives, which in turn helps them target their messages in a world already cluttered with advertisements.
SilverPush, a global platform for cross-screen marketing technology, developed Mirrors as a response to the industry problem of random ads placements in videos. This is negating the impact of advertisers, especially as video is projected to become the next universal format for content.
“Mirrors is the culmination of years of research to understand how users are engaging with brands when watching online video content. With Mirrors, we are creating endless contextual possibilities and are exploring every possible avenue to elevate the user experience for consumers; while helping established and emerging brands operating in APAC target their messages more effectively,” said Kartik Mehta, Chief Revenue Officer, SilverPush.
SilverPush’s Mirrors technology comes at a time where the growing consumer appetite for on-demand and multiscreen viewing is disrupting the global TV market. This growth is especially pronounced in the Asia-Pacific, where forecasted pay-TV and over-the-top (OTT) video will be worth $77.4 billion by 2021, according to marketing data analytics firm dataxu.
Referring to Mirrors as the potential ‘search engine’ of online video, Mehta explained that the platform’s technology — while being a B2B solution for now — can potentially power consumer interactions as well. Providing context within a video will help Mirrors become the de facto engine for people to search, explore and navigate online videos, he said.
"Imagine watching a video of a romantic song that features Emma Stone holding hands with a male co-star at night time. Mirrors can quickly go through all indexed videos to identify this moment and recommends La La Land to the viewer in return,” added Mehta.