Deep.ad Launches PanelBot Visual Media Analysis Tool

Content recognition platform Deep.ad has launched a data extraction tool called PanelBot, to enable market researchers to uncover what market intelligence visual media contains - such as brand names, logos, product categories, and trademarks - and where the asset is from.

Chicago-based Deep.ad provides machine learning-powered software, which automates the discovery and extraction of brand, product, and advertising information within visual media and digital content. Its new PanelBot platform combines object detection and optical character recognition (OCR) technology, to label 'brandprints' that appear within content - including paid media creative and user-generated content. The tool is designed to be able to pinpoint where an asset has come from - eg, Facebook, YouTubeTV, TikTok, Google etc - and what else is known about the content publisher. In addition to raw image, video, and gif files and URLs, PanelBot also scrapes PDFs, audio files, photographs, HTML pages, and other media.

According to the company, compared to manual annotation, PanelBot reduces labeling costs by 70% and lead time by 99%. CEO Kristopher Kubicki (pictured) comments: 'PanelBot is the automated labeling platform market research companies have been waiting for. It's the fastest way to go from a black hole of unstructured and 'unanalyzed' rich media to objective and actionable market intelligence.'

Web site: www.deep.ad .