Combine to Describe: Evaluating Compositional Generalization in Image Captioning.

George Pantazopoulos Alessandro Suglia Arash Eshghi

Published in: ACL (student) (2022)

Keyphrases

single image
image data
image features
multiscale
input image
image representation
image noise
image content
image analysis
template matching
high resolution
spatial information
image collections
image pixels
feature points
low level
image retrieval
edge detection
multi view
remote sensing
image matching
image set
pixel level
grey level