Fetching the paper…

Show, Attend and Tell: Neural Image Caption Generation with Visual Attention · Around