PyramidStyler: Transformer-Based Neural Style Transfer with Pyramidal Positional Encoding and Reinforcement Learning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911188927905792 |
|---|---|
| author | Durairaju, Raahul Krishna Saruladha, K. |
| author_facet | Durairaju, Raahul Krishna Saruladha, K. |
| contents | Neural Style Transfer (NST) has evolved from Gatys et al.'s (2015) CNN-based algorithm, enabling AI-driven artistic image synthesis. However, existing CNN and transformer-based models struggle to scale efficiently to complex styles and high-resolution inputs. We introduce PyramidStyler, a transformer framework with Pyramidal Positional Encoding (PPE): a hierarchical, multi-scale encoding that captures both local details and global context while reducing computational load. We further incorporate reinforcement learning to dynamically optimize stylization, accelerating convergence. Trained on Microsoft COCO and WikiArt, PyramidStyler reduces content loss by 62.6% (to 2.07) and style loss by 57.4% (to 0.86) after 4000 epochs--achieving 1.39 s inference--and yields further improvements (content 2.03; style 0.75) with minimal speed penalty (1.40 s) when using RL. These results demonstrate real-time, high-quality artistic rendering, with broad applications in media and design. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_01715 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PyramidStyler: Transformer-Based Neural Style Transfer with Pyramidal Positional Encoding and Reinforcement Learning Durairaju, Raahul Krishna Saruladha, K. Computer Vision and Pattern Recognition Artificial Intelligence Neural Style Transfer (NST) has evolved from Gatys et al.'s (2015) CNN-based algorithm, enabling AI-driven artistic image synthesis. However, existing CNN and transformer-based models struggle to scale efficiently to complex styles and high-resolution inputs. We introduce PyramidStyler, a transformer framework with Pyramidal Positional Encoding (PPE): a hierarchical, multi-scale encoding that captures both local details and global context while reducing computational load. We further incorporate reinforcement learning to dynamically optimize stylization, accelerating convergence. Trained on Microsoft COCO and WikiArt, PyramidStyler reduces content loss by 62.6% (to 2.07) and style loss by 57.4% (to 0.86) after 4000 epochs--achieving 1.39 s inference--and yields further improvements (content 2.03; style 0.75) with minimal speed penalty (1.40 s) when using RL. These results demonstrate real-time, high-quality artistic rendering, with broad applications in media and design. |
| title | PyramidStyler: Transformer-Based Neural Style Transfer with Pyramidal Positional Encoding and Reinforcement Learning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2510.01715 |