Two of our institute’s colleagues published a paper on hidden tobacco advertising in the February Frontiers in Artificial Intelligence issue. Dr. Tamás Joó, Vice President of MEMT, and his co-authors’ research used artificial intelligence to create a text and image processing model that can detect even hidden tobacco advertising in media content. The system developed by the working group achieved 74% accuracy in recognizing smoking-related images and 98% accuracy in recognizing text. The practical benefit is that we can use artificial intelligence to detect covert tobacco advertisements efficiently in the future.
Read the full article here.