CRVD
A video denoising dataset aimed at real sensor noise. It uses specialized equipment to capture noisy videos and clean references synchronously in real scenes, with noise originating from the physical characteristics of the sensor rather than added Gaussian noise. It is an important benchmark for evaluating the practical applicability of denoising methods.
Timeline (last 90 days)
FilterFusion achieves 34.72 dB PSNR and 0.9330 SSIM on the CRVD dataset.
CRVD is a video denoising dataset oriented toward real sensor noise, capturing noisy videos and clean references in real scenes synchronously through professional equipment.
FilterFusion is currently validated mainly on the CRVD dataset, and whether the filter bank's performance remains robust in more complex real-world scenarios still needs further observation.
All Facts (3)
FilterFusion achieves 34.72 dB PSNR and 0.9330 SSIM on the CRVD dataset.
50%UnverifiedCRVD is a video denoising dataset oriented toward real sensor noise, capturing noisy videos and clean references in real scenes synchronously through professional equipment.
50%UnverifiedFilterFusion is currently validated mainly on the CRVD dataset, and whether the filter bank's performance remains robust in more complex real-world scenarios still needs further observation.
50%