ErA: Error-Aware Deep Unrolling Network for Single Image Defocus Deblurring
For researchers in image deblurring, ErA provides a new state-of-the-art method that generalizes well across multiple benchmarks.
ErA achieves state-of-the-art defocus deblurring by jointly learning a compact kernel basis and per-pixel weights, with an error-aware term that corrects kernel estimation errors, reaching top PSNR/SSIM on DPDD, RealDOF, and RTF datasets.
We introduce ErA (Error-Aware Deep Unrolling Network), an end-to-end frame work for single-image defocus deblurring. ErA jointly learns a compact kerne basis and per-pixel weights, while an error-aware term in Augmented Lagrangian unrolling corrects kernel estimation errors via alternating updates and ResUNet denoisers. It achieves state-of-the-art PSNR/SSIM on DPDD, RealDOF, and RTF, and shows strong generalization on CUHK without ground truth.