mScanIt is an AI driven machine learning based solution developed by mFilterIt, the ad-fraud specialists, for the next generation of advertising - video. As per the recent Wyzowl's 'The State of Video Marketing 2019' report, 91% of the marketers consider video as a part of their marketing strategy and 99% will continue using video for marketing. At the same time 88% will spend more on video in 2019.
India is catching up the trend ahead of several global destinations. The recently released Nokia MBiT Index pegs daily data average use per user at 10GB. Out of this close to 70% of the data is consumed by video.
Foreseeing the challenges that the marketers are going to face as they engage more with video content, mFilterIt had been working on the research and development of a product that could solve many pertinent issues inherent with the present digital video content. After rigorous product development and tests and verifications in real time with few of our beta users, mScanIt was added to the stack of mFilterIt solutions for a comprehensive ad-fraud elimination.
Using proprietary algorithms developed by mFilterIt engineers, mScanIt rips apart video content and scans every frame to go through the contents. Leveraging AEye, mScanIt virtually views the video as any human and analyses it accordingly as per preset parameters.
mScanIt operates in a 3-tier approach. At Level 1, the algorithm discovers the video and collects the meta data and other descriptors. A similar analysis is done by the tool of the environment where the ad is placed. This would be a YouTube channel, OTT content or any GDN.
The second level is more sophisticated where a frame by frame analysis of the video is performed by AEye, mScanIts virtual imagery sensor. This layer adds a rigorous secure verification process which makes sure that the actual content syncs with the metadata gathered. Else, videos where the metadata is compromised or inaccurate can pass through the first iteration of fraud detection.
The final stage of validation is done by a human eye. Our video fraud analysts randomly review the video content, metadata and descriptors to ensure that the final analysis about the video is aligned and has no discrepancy.