Paper: Visual Features for Linguists: Basic image analysis techniques for multimodally-curious NLPers

ACL ID P13-5001
Title Visual Features for Linguists: Basic image analysis techniques for multimodally-curious NLPers
Venue Annual Meeting of the Association of Computational Linguistics
Session Tutorial Abstracts
Year 2013
Authors

are used to induce more abstract features, focus- ing on the well-established bags-of-visual-words method to represent images, but also briefly in- troducing more recent developments, that include capturing spatial information with pyramid repre- sentations, soft visual word clustering via Fisher encoding and attribute-based image representa- tion. Next, we will discuss some example appli- cations, and we will conclude with a brief practi- cal illustration of visual feature extraction using a software package we developed. The tutorial is addressed to computational lin-