“Dr. Fei-Fei Li's lab at Stanford, along with Google, produced the first significant work on image captioning by combining sequential models (LSTMs) and CNNs around 2013-2014.”