Denoising Large-Scale Image Captioning from Alt-text Data using Content Selection Models

Open in new window