A semi-supervised feature-level attribute manipulation method for fashion images that matches manipulated feature distributions to real features to support both instance retrieval and attribute editing.
Aggregated residual transformations for deep neural networks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Semi-supervised Feature-Level Attribute Manipulation for Fashion Image Retrieval
A semi-supervised feature-level attribute manipulation method for fashion images that matches manipulated feature distributions to real features to support both instance retrieval and attribute editing.