Deep Learning for Atmospheric Turbulence Mitigation: A Comparison Between GANs and Transformers

Key points

• Atmospheric turbulence creates complex spatial and temporal distortions that significantly reduce image quality.1
• Our results show an important finding: GAN-based models perform better in severely degraded, low-quality situations, while Transformer-based models achieve better results in moderate to high-quality settings, especially in correcting geometric distortions.1
• We also enhance the Turbulence Mitigation Transformer (TMT) by introducing a refined training strategy, an augmented input design that includes a temporally averaged frame as prior structural information, and a modified loss function that incorporates the Structural Similarity Index (SSIM).1
• Overall, this study provides an in-depth evaluation of deep learning methods for turbulence reduction and offers practical guidance for selecting models in different imaging environments.1

Sources (1)

• [1]Deep Learning for Atmospheric Turbulence Mitigation: A Comparison Between GANs and Transformers
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 10:50 PM

  This work compares two deep learning methods for turbulence reduction: Generative Adversarial Networks (GANs) and Transformer-based models.

  Atmospheric turbulence creates complex spatial and temporal distortions that significantly reduce image quality.

Extractive summary: sentences quoted from the sources.