CSHNet: A Novel Information Asymmetric Image Translation Method

TL;DR

CSHNet introduces a hybrid CNN-Swin transformer architecture with interactive and loss-based modules to improve asymmetric cross-domain image translation, especially in structure preservation and detail enhancement.

Contribution

The paper proposes a novel CNN-Swin hybrid network with interactive and edge-preserving components for improved asymmetric image translation.

Findings

Outperforms existing methods in visual quality and metrics

Effective in both scene-level and instance-level datasets

Enhances structural integrity during translation

Abstract

Despite advancements in cross-domain image translation, challenges persist in asymmetric tasks such as SAR-to-Optical and Sketch-to-Instance conversions, which involve transforming data from a less detailed domain into one with richer content. Traditional CNN-based methods are effective at capturing fine details but struggle with global structure, leading to unwanted merging of image regions. To address this, we propose the CNN-Swin Hybrid Network (CSHNet), which combines two key modules: Swin Embedded CNN (SEC) and CNN Embedded Swin (CES), forming the SEC-CES-Bottleneck (SCB). SEC leverages CNN's detailed feature extraction while integrating the Swin Transformer's structural bias. CES, in turn, preserves the Swin Transformer's global integrity, compensating for CNN's lack of focus on structure. Additionally, CSHNet includes two components designed to enhance cross-domain information retention: the Interactive Guided Connection (IGC), which enables dynamic information exchange between SEC and CES, and Adaptive Edge Perception Loss (AEPL), which maintains structural boundaries during translation. Experimental results show that CSHNet outperforms existing methods in both visual quality and performance metrics across scene-level and instance-level datasets. Our code is available at: https://github.com/XduShi/CSHNet.