Learning Descriptive Image Captioning via Semipermeable Maximum Likelihood Estimation

Open in new window