Defocus deblurring is a challenging task due to the spatially varying nature of defocus blur with multiple plausible solutions of a single given image. However, most existing methods falter when faced with extensive and variable defocus blur, either ignoring it or relying on additional loss functions to enhance perceptual quality. This often results in unrealistic reconstructions and compromised generalizability. In this paper, we propose a novel Residual Diffusion Deblurring Model framework for single image defocus deblurring. Our approach integrates a pre-trained defocus map estimator and a lightweight pre-deblur module with a learnable receptive field, providing crucial posterior information to effectively address large-scale and varying shaped defocus blur. In addition, a carefully-design denoising network enables the generation of diverse reconstructions from a single input. This approach not only significantly improves the perceptual quality of defocus deblurring outputs through multi-step residual learning, but also offers a more efficient inference strategy. Experimental results demonstrate that our method achieves competitive performance on real-world defocus deblurring image datasets across both perceptual and distortion evaluation metrics.

Synthetic aperture radar (SAR) images are pivotal in remote sensing applications. However, due to the physical characteristics of coherent imaging, current SAR image compression methods are susceptible to speckle noise, leading to distortion and higher compression rates. To address these issues, we propose a content-adaptive transformation network that dynamically adjusts the receptive field size based on image content, thereby mitigating noise impact and capturing detailed features more effectively. In addition, we developed a context-aware entropy model (CAEM) to better explore channel correlations within the latent feature space, which helps to reduce redundancy in latent features. Experimental results demonstrate that our method achieves state-of-the-art performance compared to traditional image compression standards and deep learning-based models, significantly enhancing the compression ratio and image reconstruction quality for the Sandia and ICEYE datasets.