Paper: VSEM: An open library for visual semantics representation

ACL ID P13-4032
Title VSEM: An open library for visual semantics representation
Venue Annual Meeting of the Association of Computational Linguistics
Session System Demonstration
Year 2013
Authors

VSEM is an open library for visual se- mantics. Starting from a collection of tagged images, it is possible to auto- matically construct an image-based rep- resentation of concepts by using off-the- shelf VSEM functionalities. VSEM is en- tirely written in MATLAB and its object- oriented design allows a large flexibility and reusability. The software is accompa- nied by a website with supporting docu- mentation and examples.