Ganetic loss for generative adversarial networks with a focus on medical applications : научное издание

Akhmedova Shakhnaz; Körber Nils

Дополнительные ссылки: eLibrary

База данных: Каталог библиотеки СФУ

Библиографическое описание: Ganetic loss for generative adversarial networks with a focus on medical applications : научное издание / S. Akhmedova, N. Körber. - Текст : непосредственный // Neural Computing & Applications. - 2025. - Т. 37, № 23. - P. 19177-19197. - ISSN 0941-0643. - ISSN 1433-3058, DOI 10.1007/s00521-025-11384-2.

Аннотация: Abstract Generative adversarial networks (GANs) are machine learning models that are used to estimate the underlying statistical structure of a given dataset and, as a result, can be used for a variety of tasks, such as image generation or anomaly detection. Despite their initial simplicity, designing an effective loss function for training GANs remains challenging, and various loss functions have been proposed, aiming to improve the performance and stability of generative models. In this study, loss function design for GANs is presented as an optimization problem solved using the Genetic Programming approach. Initial experiments were carried out using a small Deep Convolutional GAN model and the MNIST dataset in order to search experimentally for an improved loss function. The functions found were evaluated on CIFAR-10, with the best function, named GANetic loss, showing better performance and stability compared to the losses commonly used for GAN training. To further evaluate its general applicability to more challenging problems, GANetic loss was applied to two medical applications: image generation and anomaly detection. Experiments were performed with histopathological, gastrointestinal, and glaucoma images to evaluate the GANetic loss in medical image generation, resulting in improved image quality compared to the baseline models. The GANetic loss used for polyp and glaucoma images showed a strong improvement in the detection of anomalies. In summary, the GANetic loss function was evaluated on multiple datasets and applications, where it consistently outperformed alternative losses. Moreover, GANetic loss leads to stable training and reproducible results, a known weak spot of GANs. Code: https://github.com/ZKI-PH-ImageAnalysis/GANetic-Loss.

Год издания: 2025

Авторы: Akhmedova Shakhnaz, Körber Nils

Издательство: Springer-Verlag London Ltd

Источник: Neural Computing & Applications

Выпуск: Т. 37, № 23

Номера страниц: 19177-19197

Количество экземпляров:

Твёрдая копия издания отсутствует в фонде библиотеки

Показать дополнительные сведения

Generative adversarial network for classification of mechanical fault diagnosis model : статья из журнала

Guo C. и др.
2025 год

Multi-objective parametric synthesis of loss functions in neural networks with evolutionary… : доклад, тезисы доклада

Morozov Eduard и др.
2024 год

Methodological aspects of the regional innovative development evaluation with focus on investment… : статья из журнала

Rouiga I.R. и др.
2016 год

On programming an application to mitigate DoS attack using OpenDaylight controller in software… : статья из журнала

Cajas Guijarro C. D. и др.
2021 год

A Study on a Probabilistic Method for Designing Artificial Neural Networks for the Formation of… : научное издание

Bukhtoyarov Vladimir V. и др.
2023 год

POSITION ADAPTATION OF CANDIDATE SOLUTIONS BASED ON THEIR SUCCESS HISTORY IN NATURE-INSPIRED… : научное издание

Akhmedova Shakhnaz и др.
2019 год

Minimizing unnecessary tax audits using multi-objective hyperparameter tuning of XGBoost with focal… : научное издание

Malashin Ivan P. и др.
2025 год

АВТОМАТИЧЕСКАЯ ГЕНЕРАЦИЯ ОБУЧАЮЩИХ НАБОРОВ ДАННЫХ С ИСПОЛЬЗОВАНИЕМ 3D-МОДЕЛЕЙ ДЛЯ ЗАДАЧИ… : статья из журнала

Чичиков С.А. и др.
2024 год

Использование нейронных сетей для обеспечения выразительности и динамики статических изображений : статья из журнала

Еремеев Д.Е. и др.
2024 год

Библиотечно-издательский комплекс СФУ

Ganetic loss for generative adversarial networks with a focus on medical applications : научное издание
статья из журнала

Похожие публикации