Year: 2017
Type: conference-paper
Abstract: Leveraging class semantic descriptions and examples of known objects, zero-shot learning makes it possible to train a recognition model for an object class whose examples are not available. In this paper, we propose a novel zero-shot learning model that takes advantage of clustering structures in the semantic embedding space. The key idea is to impose the structural constraint that semantic representations must be predictive of the locations of their corresponding visual exemplars. To this end, this reduces to training multiple kernel-based regressors from semantic representation-exemplar pairs from labeled data of the seen object categories. Despite its simplicity, our approach significantly outperforms existing zero-shot learning methods on standard benchmark datasets, including the ImageNet dataset with more than 20,000 unseen categories.
Authors: Soravit Changpinyo, Wei-Lun Chao, Fei Sha
Institution: University of Southern California
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: RI: Medium: Collaborative Research: Learning to Summarize User-Generated Video, NRI-Small: Spacial Primitives for Enabling Situated Human-Robot Interaction, RI: Medium: Collaborative Research: Learning to Su, G5385673730, G7452299184, Collaborative Research: Socially Assistive Robots, G8257659198, RI: Medium: Collaborative Research: Semantically Discriminative: Guiding Mid-Level Representations for Visual Object Recognition with External Knowledge, EAGER: Leveraging Structure to Realize the Promise of Transfer Learning
OpenAlex ID: W2962762077
