Image Captioning through Image Transformer

Sen He, Wentong Liao, Hamed R. Tavakoli, M.Y. Yang, Bodo Rosenhahn, Nicolas Pugeault

Research output: Working paperPreprintAcademic

179 Downloads (Pure)

Abstract

Automatic captioning of images is a task that combines the challenges of image analysis and text generation. One important aspect in captioning is the notion of attention: How to decide what to describe and in which order. Inspired by the successes in text analysis and translation, previous work have proposed the \textit{transformer} architecture for image captioning. However, the structure between the \textit{semantic units} in images (usually the detected regions from object detection model) and sentences (each single word) is different. Limited work has been done to adapt the transformer's internal architecture to images. In this work, we introduce the \textbf{\textit{image transformer}}, which consists of a modified encoding transformer and an implicit decoding transformer, motivated by the relative spatial relationship between image regions. Our design widen the original transformer layer's inner architecture to adapt to the structure of images. With only regions feature as inputs, our model achieves new state-of-the-art performance on both MSCOCO offline and online testing benchmarks.
Original languageEnglish
PublisherArXiv.org
Pages1-17
Number of pages17
DOIs
Publication statusPublished - 29 Apr 2020

Keywords

  • cs.CV

Fingerprint

Dive into the research topics of 'Image Captioning through Image Transformer'. Together they form a unique fingerprint.

Cite this