• Home
  • Search
  • Structure Preserving Convolutional Attention for Image Captioning
  • Open Access IconOpen Access
  • Cite Icon2
  • https://doi.org/10.3390/app9142888Copy DOI Icon

Structure Preserving Convolutional Attention for Image Captioning

Show More
  • Abstract
  • Highlights & Summary
  • PDF
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

In the task of image captioning, learning the attentive image regions is necessary to adaptively and precisely focus on the object semantics relevant to each decoded word. In this paper, we propose a convolutional attention module that can preserve the spatial structure of the image by performing the convolution operation directly on the 2D feature maps. The proposed attention mechanism contains two components: convolutional spatial attention and cross-channel attention, aiming to determine the intended regions to describe the image along the spatial and channel dimensions, respectively. Both of the two attentions are calculated at each decoding step. In order to preserve the spatial structure, instead of operating on the vector representation of each image grid, the two attention components are both computed directly on the entire feature maps with convolution operations. Experiments on two large-scale datasets (MSCOCO and Flickr30K) demonstrate the outstanding performance of our proposed method.

Loading PDF

Similar Papers
  • Conference Article

Text generation image method based on convolutional attention and cross-domain feature fusion

  • Sep 22, 2023
  • Pingan Qiao +2
  • Conference Article
  • Citations2272

SCA-CNN: Spatial and Channel-Wise Attention in Convolutional Networks for Image Captioning

  • Jul 01, 2017
  • Long Chen +6
  • Research Article
  • Citations12

Improved skeleton-based activity recognition using convolutional block attention module

  • Apr 04, 2024
  • Computers and Electrical Engineering
  • Jing Qin +5
  • PDF
  • Research Article
  • Citations11

AttentNet: Fully Convolutional 3D Attention for Lung Nodule Detection

  • Mar 16, 2025
  • SN Computer Science
  • Majedaldein Almahasneh +2
  • Research Article

Rail State Recognition Method Based on a Convolutional Block Attention Module

  • Oct 01, 2024
  • Journal of Physics: Conference Series
  • Lili Liu +3
  • Book Chapter
  • Citations21

Iconographic Image Captioning for Artworks

  • Jan 01, 2021
  • Eva Cetinic
  • Research Article
  • Citations6

Eye-movement-prompted large image captioning model

  • Nov 01, 2024
  • Pattern Recognition
  • Zheng Yang +3
  • Research Article
  • Citations165

Context-Aware Visual Policy Network for Fine-Grained Image Captioning

  • Apr 09, 2019
  • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • Zheng-Jun Zha +4
  • Research Article
  • Citations29

Hybrid attention network for image captioning

  • May 13, 2022
  • Displays
  • Wenhui Jiang +4
  • Conference Article

Research on Automatic Extraction of Tobacco Plants Based on U2-Net of Convolutional Block Attention Module and UAV Visible Images

  • Aug 19, 2022
  • Chao Feng +5
  • Research Article
  • Citations1

ACU-TransNet: Attention and convolution-augmented UNet-transformer network for polyp segmentation.

  • Oct 12, 2024
  • Journal of X-ray science and technology
  • Lei Huang +1
  • Conference Article
  • Citations2

Research on Image Caption Based on Multiple Word Embedding Representations

  • Mar 01, 2021
  • Zhen-Xian Lin +4
  • Research Article
  • Citations3

A Survey on Image Captioning datasets and Evaluation Metrics

  • Apr 01, 2021
  • IOP Conference Series: Materials Science and Engineering
  • Himanshu Sharma
  • Research Article

From Pixels to Sentence: A Comprehensive Study of Transformers-Based Models for Image Captioning

  • Mar 30, 2026
  • Journal of Al-Qadisiyah for Computer Science and Mathematics
  • Haider Jaber Samawi +1
  • Conference Article

SAM-MFF: An Image Captioning Generation Method Based on Multi-layer Feature Fusion

  • Sep 20, 2024
  • Ziqing He +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.