Packaging branding reduction design based on VGG-CycleGAN model incorporating sustainable environmental visual elements

Read full text Summarise with VIRAT
Authors: Jiaxuan Li, Peisheng Yu
Cited by: 0
References: 19

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The amelioration of individuals’ living standards persists, concomitantly fostering an accelerated evolution within the packaging industry. Concurrently, the ethos of sustainability deepens its roots within the collective consciousness. A discernible trend emerges wherein an increasing number of designs underscore their ornamental facade, neglecting practicality, user-friendliness, humaneness, and environmental stewardship. Consequently, the investigation into sustainable packaging reduction emerges as a pivotal research focus at the intersection of packaging design and environmental conservation. In this treatise, we propose a framework for packaging design within the context of sustainable and verdant progress—a generative style migration network. This innovative framework facilitates the evolution of packaging styles and accelerates swift design processes, thereby streamlining the lightweight design cycle of packaging. The foundational augmentation involves incorporating a Visual Geometry Group (VGG) feature comparison network into the loss function comparison segment of CycleGAN. Crucially, constraints are appended to ensure the model’s stability. Throughout the model’s training phase, a comprehensive comparative analysis unfolds, encompassing datasets of cartoon, sketch, and authentic packaging images. Objective quantitative indices, namely FID, SSIM, and PSNR, serve as benchmarks. The empirical results affirm the commendable performance of the proposed framework. In the subjective realm, satisfaction evaluations corroborate the framework’s efficacy, maintaining a notable 50% satisfaction rate across diverse style comparisons. This substantiates the framework’s image generation prowess, providing a robust technical foundation for the future landscape of multi-stylistic packaging design.

Fields of Study & Subjects

Generative Adversarial Networks and Image Synthesis Advanced Neural Network Applications Digital Media and Visual Art

Publication details

Journal / source
PeerJ Computer Science
Volume
12
Pages
e3683–e3683
Publication date
25 March 2026
Language
EN