About the Turbulence Simulator

Due to the difficulties in collecting distorted/clean image pairs, researchers in the field of atmospheric turbulence mitigation are mostly working with synthetic data. The P2S simulator is the latest atmospheric turbulence simulator developed by Purdue University. It is, to our knowledge, the first physical-based simulator that can achieve close to real-time speed for simulating the atmospheric turbulence effect on images.

Paper and simulator code: paper, code

Evaluation

In the final testing stage, there will be 50 sequences of hot-air images and 500 sequences of text patterns. Each sequence will contain 100 turbulence distorted frames. The task for the participants is to reconstruct a high quality image from these 100 distorted frames (you can use all or part of these frames). The final ranking of the challenge will be based on the average accuracy of three existing scene text recognition algorithms on the reconstruction result of the text patterns. To encourage the participants to develop methods for generic scene, the same model used for processing the text images must also achieve certain performance (PSNR) on the hot-air images. The threshold is to be set by the organizer.

• Download: Starting Kit (including evaluation tools, dry run dataset, and a description of the utilities)
• Download NEW!: Final Testing Data
• Latest Release! Please find the complete dataset at: Our site.
• Codalab: Codalab Link
• Simulator Code: Reference Github Repository

If you have any questions about this challenge track please feel free to email cvpr2022.ug2challenge@gmail.com