Enhancing Visual Embeddings through Weakly Supervised Captioning for Zero-Shot Learning.

Matteo Bustreo Jacopo Cavazza Vittorio Murino

Published in: ICCV Workshops (2019)

Keyphrases

weakly supervised
topic models
object class
superpixels
relation extraction
visual features
low level
named entities
semi supervised
viewpoint
domain specific
image patches