Employing Hierarchical Clustering and Reinforcement Learning for Attribute-Based Zero-Shot Classification

Published: 15 May 2019| Version 1 | DOI: 10.17632/r3v39665bm.1
Contributor:
Bin Liu

Description

low-level features extracted from AwA and aPaY using vgg19 and googlenet and attribute annotations of labels; HCRL implemented on Weka; Configuration of the experiments in the corresponding article.

Files

Steps to reproduce

The original image file is too large, please refer to the footnote cited in the article URL. Training data used 5% and 3% of the sample size respectively.

Institutions

National University of Defense Technology

Categories

Machine Learning

Licence