AesFA：一种美学特征感知的任意神经风格迁移
AesFA: An Aesthetic Feature-Aware Arbitrary Neural Style Transfer
AI总结：
本文提出轻量级美学特征感知神经风格迁移模型AesFA，通过频率分解解耦美学风格并以端到端方式训练，引入对比损失提升风格化质量，在质量和推理速度上均优于现有方法。
摘要：
神经风格迁移（NST）近年来取得了显著发展。然而，尽管进展迅速，现有NST方法要么难以有效传递风格中的美学信息，要么由于使用预训练模型而导致计算成本高且特征解耦效率低下。本文提出了一种轻量但有效的模型——AesFA，即美学特征感知的神经风格迁移。其核心思想是通过图像频率对图像进行分解，从而更好地从参考图像中解耦美学风格，同时以端到端方式训练整个模型，在推理阶段完全排除预训练模型。为了提高网络提取更具区分性表示的能力并进一步提升风格化质量，本文引入了一种新的美学特征：对比损失。大量实验和消融研究表明，该方法不仅在风格化质量上优于近期的NST方法，而且实现了更快的推理速度。代码可在 https://github.com/Sooyyoungg/AesFA 获取。
英文摘要：
Neural style transfer (NST) has evolved significantly in recent years. Yet, despite its rapid progress and advancement, existing NST methods either struggle to transfer aesthetic information from a style effectively or suffer from high computational costs and inefficiencies in feature disentanglement due to using pre-trained models. This work proposes a lightweight but effective model, AesFA -- Aesthetic Feature-Aware NST. The primary idea is to decompose the image via its frequencies to better disentangle aesthetic styles from the reference image while training the entire model in an end-to-end manner to exclude pre-trained models at inference completely. To improve the network's ability to extract more distinct representations and further enhance the stylization quality, this work introduces a new aesthetic feature: contrastive loss. Extensive experiments and ablations show the approach not only outperforms recent NST methods in terms of stylization quality, but it also achieves faster inference. Codes are available at https://github.com/Sooyyoungg/AesFA.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
