DeepShoe: An improved Multi-Task View-invariant CNN for street-to-shop shoe retrieval.

Huijing Zhan Boxin Shi Ling-Yu Duan Alex C. Kot

Published in: Comput. Vis. Image Underst. (2019)

Keyphrases

multi task
view invariant
multi task learning
learning tasks
multi class
information retrieval

feature selection
multiple tasks
learning problems
sparse learning
action recognition

multi view
transfer learning
single view
gaussian processes
human motion

image retrieval
support vector
image processing
learning algorithm
data mining