Contextualize, Show and Tell: A Neural Visual Storyteller

TL;DR

This paper introduces a neural model that generates coherent short stories from image sequences by extending image description techniques with context-aware LSTM encoders and decoders, achieving competitive results in storytelling benchmarks.

Contribution

It proposes a novel neural architecture that models context across image sequences for storytelling, improving upon previous image description models.

Findings

Achieved competitive METEOR scores

Received high human ratings in storytelling quality

Demonstrated effective context modeling across images

Abstract

We present a neural model for generating short stories from image sequences, which extends the image description model by Vinyals et al. (Vinyals et al., 2015). This extension relies on an encoder LSTM to compute a context vector of each story from the image sequence. This context vector is used as the first state of multiple independent decoder LSTMs, each of which generates the portion of the story corresponding to each image in the sequence by taking the image embedding as the first input. Our model showed competitive results with the METEOR metric and human ratings in the internal track of the Visual Storytelling Challenge 2018.