From Images to Sentences through Scene Description Graphs using Commonsense Reasoning and Knowledge

TL;DR

This paper introduces a method for generating detailed image descriptions by constructing scene description graphs using vision, reasoning, and knowledge bases, improving relevance over existing captioning methods.

Contribution

The paper presents a novel approach combining vision, commonsense reasoning, and lexical knowledge to generate more comprehensive image descriptions via scene description graphs.

Findings

Generated descriptions are more relevant and thorough than recent state-of-the-art captioning methods.

Evaluation shows comparable image-sentence alignment performance to current leading approaches.

Method outperforms existing models on multiple datasets in relevance and detail.

Abstract

In this paper we propose the construction of linguistic descriptions of images. This is achieved through the extraction of scene description graphs (SDGs) from visual scenes using an automatically constructed knowledge base. SDGs are constructed using both vision and reasoning. Specifically, commonsense reasoning is applied on (a) detections obtained from existing perception methods on given images, (b) a "commonsense" knowledge base constructed using natural language processing of image annotations and (c) lexical ontological knowledge from resources such as WordNet. Amazon Mechanical Turk(AMT)-based evaluations on Flickr8k, Flickr30k and MS-COCO datasets show that in most cases, sentences auto-constructed from SDGs obtained by our method give a more relevant and thorough description of an image than a recent state-of-the-art image caption based approach. Our Image-Sentence Alignment Evaluation results are also comparable to that of the recent state-of-the art approaches.