Presentation Information

1:45 PM - 2:00 PM JST(4:45 AM - 5:00 AM UTC)Candidate for Academic Incentive Award

[A-8-15]Consideration of Missing Pixel Restoration with Neural Ordinary Differential Equations

◎Takuro Arita¹, Michiharu Maeda¹ (1. Fukuoka Institute of Technology)

Keywords:

Neural Ordinary Differential Equations,Missing Pixel Restoration,Interpolation

This study investigates a missing pixel restoration with neural ordinary differential equations (Neural ODEs). A model is constructed to estimate pixel values from image coordinates and applied to the restoration of images containing randomly missing pixels. Comparative experiments are conducted against median filtering and smoothing filtering methods, and the performance is evaluated using peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM).

Back to Session information