Caption This: Artificial intelligence aids visual storytelling
If a picture is worth 1,000 words, then social media users are talking a lot. On average, more than 60 million photos are uploaded to Instagram every day. On Facebook, that daily average reaches at least 350 million. Every photo contains a narrative frozen in time, making captions critical to interpreting and connecting to the scene.
Francis Ferraro, a PhD candidate in the Johns Hopkins Department of Computer Science and the Johns Hopkins Center for Language and Speech Processing (CLSP), has been working with Microsoft Research to develop an automatic image captioning system. Their goal is clear: develop a system that can describe an image with context and an emotional connection, the same way a human can.
The project is led by Margaret Mitchell, a former postdoctoral scholar in the Johns Hopkins Human Language Technology Center of Excellence (HLTCOE). She worked under Benjamin Van Durme, assistant professor of computer science and a CLSP affiliate.