Guiding Long-Short Term Memory for Image Caption Generation

TL;DR

This paper introduces gLSTM, an enhanced LSTM model for image captioning that incorporates semantic image information and improved beam search strategies, achieving competitive results on standard datasets.

Contribution

The paper presents a novel gLSTM model that integrates semantic image features into each LSTM unit for better caption generation.

Findings

gLSTM outperforms standard LSTM on benchmark datasets

Semantic guidance improves caption relevance

Length normalization enhances beam search results

Abstract

In this work we focus on the problem of image caption generation. We propose an extension of the long short term memory (LSTM) model, which we coin gLSTM for short. In particular, we add semantic information extracted from the image as extra input to each unit of the LSTM block, with the aim of guiding the model towards solutions that are more tightly coupled to the image content. Additionally, we explore different length normalization strategies for beam search in order to prevent from favoring short sentences. On various benchmark datasets such as Flickr8K, Flickr30K and MS COCO, we obtain results that are on par with or even outperform the current state-of-the-art.