basicsr.models.realesrgan_model

class basicsr.models.realesrgan_model.RealESRGANModel(opt)[source]

Bases: SRGANModel

RealESRGAN Model for Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data.

It mainly performs: 1. randomly synthesize LQ images in GPU tensors 2. optimize the networks with GAN training.

feed_data(data)[source]

Accept data from dataloader, and then add two-order degradations to obtain LQ images.

nondist_validation(dataloader, current_iter, tb_logger, save_img)[source]
optimize_parameters(current_iter)[source]