Real-World Deep Local Motion Deblurring

Most existing deblurring methods focus on removing global blur caused by camera shake, while they cannot well handle local object movements. To fill the vacancy of in real scenes, we establish first motion dataset (ReLoBlur), which is captured a synchronized beam-splitting photographing system and corrected post-progressing pipeline. Based ReLoBlur, propose Local Blur-Aware Gated network (LBAG) several blur-aware techniques to bridge gap between deblurring: 1) detection approach based background subtraction localize blurred regions; 2) gate mechanism guide our 3) patch cropping strategy address data imbalance problem. Extensive experiments prove reliability ReLoBlur dataset, demonstrate that LBAG achieves better performance than state-of-the-art proposed are effective.