3 ms·Show, Attend and Tell: Neural Image Caption Generation with Visual Attention1 points by mayava 9y ago