Creativity: Generating Diverse Questions using Variational Autoencoders

TL;DR

This paper presents a novel method combining variational autoencoders and LSTMs to generate diverse, plausible questions from images, enhancing AI's creative question generation capabilities.

Contribution

It introduces a creative algorithm that leverages VAEs and LSTMs to produce a wide variety of questions from a single image, addressing the need for diversity.

Findings

Able to generate a large set of varying questions from one image

Demonstrates the effectiveness of combining VAEs with LSTMs for diversity

Enhances AI's ability to produce creative and plausible questions

Abstract

Generating diverse questions for given images is an important task for computational education, entertainment and AI assistants. Different from many conventional prediction techniques is the need for algorithms to generate a diverse set of plausible questions, which we refer to as "creativity". In this paper we propose a creative algorithm for visual question generation which combines the advantages of variational autoencoders with long short-term memory networks. We demonstrate that our framework is able to generate a large set of varying questions given a single input image.