Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
Focusing attention at the level of objects and other salient image regions.
2017
2016
ECCV 2016
What’s missing from deep learning?
SPICE: Semantic Propositional Image Caption Evaluation
Evaluating image captions using scene graph tuples.
2015
CVPR Highlights 2015
Picking the low-hanging deep fruit.
Cheating Chess Players and Moravec’s Paradox
If an iPhone can threaten a Grandmaster, why can’t I get a robot to unpack the dishwasher?