Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 04020 | |
Number of page(s) | 6 | |
Section | AI and Advanced Applications | |
DOI | https://doi.org/10.1051/itmconf/20257004020 | |
Published online | 23 January 2025 |
DragGAN: Interactive Point-Based Image Manipulation on Generative Adversarial Networks
Data Science Institute, Shanghai Lida University, Shanghai, 201608, China
Corresponding author: yuwangbao@ldy.edu.rs
Users have increasingly demanded greater control over generated images, including flexibility, precision, and versatility as a result of the development of Generative Adversarial Networks (GANs). This post introduces DragGAN, a picture-enhancing method that uses engaging dragging to obtain exact control over philosophical photos. DragGAN enables users to move, and adjust the position, style, location, and size of specific areas of a picture, and do so by integrating feature-operated motion supervision and point-tracking techniques. DragGAN demonstrates its ability to move key points more specifically to the precise positions in the framework of combined image restoration, for instance. Experimental results demonstrate that DragGAN outperforms conventional methods in terms of the generated graphics ‘realism and objective-level accuracy. This method significantly enhances the flexibility and efficiency of picture editing, lowers the technical barrier, and enables quasi-expert users to easily accomplish higher-quality image editing, marking a significant advancement in the field of image synthesis. Future research will focus on curbing reliance on write-up-trained GAN patterns and increasing the person’s steadiness and accuracy in complex scenes. This indicates that DragGAN’s engineering is still developing, and future additions and changes may be made to improve the user experience and control results.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.