Search papers, labs, and topics across Lattice.
This paper introduces the RealDefocus Benchmark, a comprehensive framework for evaluating Single-Image Defocus Deblurring (SIDD) by providing high-resolution, well-aligned pairs of defocused and sharp images. The benchmark addresses the critical challenge of reproducibility in defocus deblurring research by offering standardized training, validation, and test splits, along with a unified evaluation protocol. Key results demonstrate improved assessment of reconstruction quality and generalization across different datasets, facilitating more rigorous comparisons among image restoration and neural rendering methods.
A new benchmark reveals that existing defocus deblurring methods struggle with generalization across datasets, highlighting a critical gap in current evaluation practices.
Single-Image Defocus Deblurring (SIDD) aims to recover an all-in-focus image from a single defocused observation, but rigorous and reproducible evaluation remains challenging due to the scarcity of realistic, high-resolution datasets with well-aligned defocused/sharp pairs and standardized protocols. We build on RealDefocus, a benchmark derived from the real-world RealBokeh dataset originally proposed for Bokeh Rendering. RealDefocus provides paired defocused inputs and sharp ground truth images, predefined training/validation/test splits, and a unified evaluation framework for comparing image restoration and neural rendering approaches. We further outline a benchmarking protocol with cross-dataset validation to assess reconstruction quality and generalization. The project page is publicly available at: www.github.com/TimSeizinger/RealDefocus-Benchmark.
