CVMar 20, 2021

3M: Multi-style image caption generation using Multi-modality features under Multi-UPDOWN model

arXiv:2103.11186v17 citations
Originality Incremental advance
AI Analysis

This work addresses the problem of generating stylized captions for images, which is incremental as it builds on existing captioning methods by incorporating multi-modality features.

The paper tackles stylish image captioning by proposing the 3M model, which uses multi-modality features and a Multi-UPDOWN architecture to generate human-like captions, achieving competitive performance on datasets like PERSONALITY-CAPTIONS and FlickrStyle10K as measured by metrics such as BLEU and CIDEr.

In this paper, we build a multi-style generative model for stylish image captioning which uses multi-modality image features, ResNeXt features and text features generated by DenseCap. We propose the 3M model, a Multi-UPDOWN caption model that encodes multi-modality features and decode them to captions. We demonstrate the effectiveness of our model on generating human-like captions by examining its performance on two datasets, the PERSONALITY-CAPTIONS dataset and the FlickrStyle10K dataset. We compare against a variety of state-of-the-art baselines on various automatic NLP metrics such as BLEU, ROUGE-L, CIDEr, SPICE, etc. A qualitative study has also been done to verify our 3M model can be used for generating different stylized captions.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes