He, S., Liao, W., Tavakoli, H. R., Yang, M., Rosenhahn, B. and Pugeault, N. (2021) Image Captioning through Image Transformer. In: 15th Asian Conference on Computer Vision, 30 Nov-04 Dec 2020, pp. 153-169. ISBN 9783030695378 (doi: 10.1007/978-3-030-69538-5_10)
Text
223861.pdf - Accepted Version 2MB |
Abstract
Automatic captioning of images is a task that combines the challenges of image analysis and text generation. One important aspect of 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 works have proposed the transformer architecture for image captioning. However, the structure between the 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 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 widens 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. The code is available at https://github.com/wtliao/ImageTransformer.
Item Type: | Conference Proceedings |
---|---|
Status: | Published |
Refereed: | Yes |
Glasgow Author(s) Enlighten ID: | Pugeault, Dr Nicolas |
Authors: | He, S., Liao, W., Tavakoli, H. R., Yang, M., Rosenhahn, B., and Pugeault, N. |
College/School: | College of Science and Engineering > School of Computing Science |
ISSN: | 0302-9743 |
ISBN: | 9783030695378 |
Published Online: | 25 February 2021 |
Copyright Holders: | Copyright © 2021 Springer Nature Switzerland AG |
First Published: | First published in Computer Vision – ACCV 2020. ACCV 2020. Lecture Notes in Computer Science, vol 12625, pp 153-169 |
Publisher Policy: | Reproduced in accordance with the copyright policy of the publisher |
Related URLs: |
University Staff: Request a correction | Enlighten Editors: Update this record