From Zero-shot Learning to Conventional Supervised Classification: Unseen Visual Data Synthesis

Long, Yang and Liu, Li and Shao, Ling and Shen, Fumin and Ding, Guiguang and Han, Jungong (2017) From Zero-shot Learning to Conventional Supervised Classification: Unseen Visual Data Synthesis. In: CVPR 2017. Computer Vision Foundation, pp. 1627-1636.

[img]
Preview
PDF (Long_From_Zero-Shot_Learning_CVPR_2017_paper)
Long_From_Zero_Shot_Learning_CVPR_2017_paper.pdf - Published Version
Available under License Creative Commons Attribution.

Download (4MB)

Abstract

Robust object recognition systems usually rely on powerful feature extraction mechanisms from a large number of real images. However, in many realistic applications, collecting sufficient images for ever-growing new classes is unattainable. In this paper, we propose a new Zero-shot learning (ZSL) framework that can synthesise visual features for unseen classes without acquiring real images. Using the proposed Unseen Visual Data Synthesis (UVDS) algorithm, semantic attributes are effectively utilised as an intermediate clue to synthesise unseen visual features at the training stage. Hereafter, ZSL recognition is converted into the conventional supervised problem, i.e. the synthesised visual features can be straightforwardly fed to typical classifiers such as SVM. On four benchmark datasets, we demonstrate the benefit of using synthesised unseen data. Extensive experimental results suggest that our proposed approach significantly improve the state-of-the-art results.

Item Type:
Contribution in Book/Report/Proceedings
ID Code:
87900
Deposited By:
Deposited On:
06 Oct 2017 20:07
Refereed?:
Yes
Published?:
Published
Last Modified:
25 Sep 2020 05:38