GRIT: Faster and Better Image Captioning Transformer Using Dual Visual Features.

Van-Quang Nguyen, Masanori Suganuma, Takayuki Okatani

Browse the full ECCV paper archive.