CapEnrich: Enriching Caption Semantics for Web Images via Cross-modal Pre-trained Knowledge.

Linli Yao Weijing Chen Qin Jin

Published in: CoRR (2022)

Keyphrases

cross modal
web images
pre trained
multi modal
semantic information
visual features
visual similarity
multimedia retrieval
image search
data sets
training data
object recognition
training set
high dimensional
prior knowledge
image classification