Dynamically Typed

Open Images v6

Google released Open Images V6, a new version of “the largest annotated image dataset in many regards.” It now features local narratives, such as the one I embedded above, consisting of “synchronized voice, text, and mouse traces over the objects being described.”