Login / Signup

MRECN: mixed representation enhanced (de)compositional network for caption generation from visual features, modeling as pseudo tensor product representation.

Chiranjib Sur
Published in: Int. J. Multim. Inf. Retr. (2020)
Keyphrases
  • visual features
  • image classification
  • image representation
  • image search
  • bag of features
  • visual information
  • semantic concepts
  • feature selection
  • image retrieval
  • low level features
  • tensor product