Abstract
This study introduces an enhanced approach to video super-resolution by extending ordinary Single-Image Super-Resolution (SISR) Super-Resolution Generative Adversarial Network (SRGAN) structure to handle spatio-temporal data. While SRGAN has proven effective for single-image enhancement, its design does not account for the temporal continuity required in video processing. To address this, a modified framework that incorporates 3D Non-Local Blocks is proposed, which is enabling the model to capture relationships across both spatial and temporal dimensions. An experimental training pipeline is developed, based on patch-wise learning and advanced data degradation techniques, to simulate real-world video conditions and learn from both local and global structures and details. This helps the model generalize better and maintain stability across varying video content while maintaining the general structure besides the pixel-wise correctness. Two model variants-one larger and one more lightweight-are presented to explore the trade-offs between performance and efficiency. The results demonstrate improved temporal coherence, sharper textures, and fewer visual artifacts compared to traditional single-image methods. This work contributes to the development of practical, learning-based solutions for video enhancement tasks, with potential applications in streaming, gaming, and digital restoration.
Abstract (translated)
这项研究提出了一种改进的方法，通过将普通的单图像超分辨率（SISR）生成对抗网络（SRGAN）结构扩展到处理时空数据来增强视频超分辨率。虽然SRGAN在单一图像增强方面表现出色，但其设计并未考虑视频处理中所需的时序连续性。为解决这一问题，提出了一个修改后的框架，该框架引入了3D非局部块，使模型能够捕捉空间和时间维度上的关系。 为了模拟现实世界的视频条件并从局部和全局结构及细节中学习，开发了一种基于分片式学习和高级数据退化技术的实验性训练流程。这有助于模型更好地泛化，并在处理不同类型的视频内容时保持稳定性，同时维持整体结构以及像素级别的准确性。 本研究提出了两种模型变体——一种较大、另一种较轻量级的，以探索性能与效率之间的权衡。结果表明，在时序一致性、更清晰的纹理和较少视觉伪影方面，相比传统的单图像方法有显著改进。这项工作为视频增强任务开发实用的学习型解决方案做出了贡献，并可能在流媒体、游戏和数字修复等领域得到应用。
URL
https://arxiv.org/abs/2505.10589
