Open Access

ARTICLE

An Image Inpainting Approach Based on Parallel Dual-Branch Learnable Transformer Network

Rongrong Gong^1,#, Tingxian Zhang^2,#, Yawen Wei², Dengyong Zhang², Yan Li^3,*

1 School of Software, Changsha Social Work College, Changsha, 410004, China
2 School of Computer Science and Technology, Changsha University of Science and Technology, Changsha, 410076, China
3 Department of Computer Engineering, INHA University, Incheon, 22201, Republic of Korea

* Corresponding Author: Yan Li. Email:
# These authors contributed equally to this work

(This article belongs to the Special Issue: Omnipresent AI in the Cloud Era Reshaping Distributed Computation and Adaptive Systems for Modern Applications)

Computers, Materials & Continua 2025, 85(1), 1221-1234. https://doi.org/10.32604/cmc.2025.066842

Received 18 April 2025; Accepted 30 June 2025; Issue published 29 August 2025

Abstract

Image inpainting refers to synthesizing missing content in an image based on known information to restore occluded or damaged regions, which is a typical manifestation of this trend. With the increasing complexity of image in tasks and the growth of data scale, existing deep learning methods still have some limitations. For example, they lack the ability to capture long-range dependencies and their performance in handling multi-scale image structures is suboptimal. To solve this problem, the paper proposes an image inpainting method based on the parallel dual-branch learnable Transformer network. The encoder of the proposed model generator consists of a dual-branch parallel structure with stacked CNN blocks and Transformer blocks, aiming to extract global and local feature information from images. Furthermore, a dual-branch fusion module is adopted to combine the features obtained from both branches. Additionally, a gated full-scale skip connection module is proposed to further enhance the coherence of the inpainting results and alleviate information loss. Finally, experimental results from the three public datasets demonstrate the superior performance of the proposed method.

---

Keywords

---