Skip to content
Conference Open access

Backbone Choice Matters: Human and Perceptual Evaluation of CNN-Based Neural Style Transfer

Oct 2026 · International Conference on Machine Learning and Applications · pp. 67-72 · 0 citations · 24 references

TL;DR

An empirical study of five CNN backbones for NST across multiple content images and artistic styles suggests that CNN backbone selection plays an important role in NST output quality and that human-centered evaluation remains necessary for assessing artistic style transfers.

Abstract

Neural style transfer (NST) commonly relies on pretrained convolutional neural networks (CNNs) to extract content and style representations, yet the effect of CNN backbone selection on the perceived quality of stylized images remains underexplored. This paper presents an empirical study of five CNN backbones for NST across multiple content images and artistic styles. We evaluate the generated images using both objective image quality assessment metrics and a human participant study with 96 respondents. Results show that images generated using higher-performing ImageNet backbones were more frequently preferred by participants and were often judged as better matching the target artistic style. However, standard image quality metrics only partially aligned with human preference and were weak predictors of perceived style imitation. These findings suggest that CNN backbone selection plays an important role in NST output quality and that human-centered evaluation remains necessary for assessing artistic style transfer systems.

Read PDF

Similar papers

Preprint Oct 2026

Can We Use Saliency Maps for the Evaluation of Generative Image Models?

Evaluating generative image models remains challenging, as commonly used metrics depend on fixed pretrained feature extractors optimized for different tasks that do not always align with human realism perception across datasets and model families. To address this limitation, we investigate a human-inspired evaluation f...

E. Ristori, L. Bindini, Marco Lippi et al. · 0 citations
Conference Open access Sep 2026

Explainable Multi-Cue Attention Networks for Deepfake Face Detection and Human-Aligned Decision Support

Generative face models are becoming harder to distinguish from real photos, as the face images produced by adversarial and style-based models become more realistic. To assess the efficacy of StyleGAN2 for generating human faces, a recent large-scale study of human subjects was conducted, where unaided observers correct...

R. Vidya, Selvamani S, L. Andal et al. · 0 citations
Book Open access Nov 2026

AesCAST-An Automated Framework for Aesthetically-Guided Style Transfer

Creative-media workflows such as visual-asset authoring and previsualization require a suitable reference style, yet existing systems usually leave that selection to the user and optimize primarily for feature correspondence. We introduce AesCAST, an automated framework that couples style recommendation with aesthetic-...

Jiang-Qi Ye, Can Zhong, Qian-Qian Qiao et al. · 0 citations
Conference Oct 2026

Diffusion-based image style transfer via parameter-free saliency recalibration

Image style transfer focuses on rendering a content image with the visual appearance of a reference artwork while keeping its main semantic layout unchanged. StyleID enables image-guided stylization without additional training by recording attention features during DDIM inversion, yet its results may still suffer from...

Hai-Tao Xin, Shi-Lin Bo · 0 citations
Conference Open access 2026

Cross-domain Image Generation: Style Transfer and Image-to-image Translation

. With the development of deep learning, cross-domain image generation has become an important research direction in computer vision, with broad application prospects in artistic creation and image editing. This paper summarizes the research results, principles, and applications of cross-domain image generation from tw...

Peng-Fei Ran · 0 citations
Open access Oct 2026

Subspace Image Synthesis for Style-Invariant Object Recognition

Human-like behavior in visual models does not necessarily imply human-like internal computation. When a model recognizes objects robustly across changes in visual style, what internal computations support this behavior, and which are relevant to human perception? We addressed this question by pruning style-robust neura...

Hayato Ono, Ayumu Yamashita, Kaoru Amano et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.