Blind Image Deconvolution by Generative-based Kernel Prior and Initializer via Latent Encoding
This work addresses a specific bottleneck in image processing for applications like photo restoration, but it is incremental as it builds on prior DIP-based approaches.
The paper tackles the sensitivity of deep image prior (DIP) methods to kernel initialization in blind image deconvolution by proposing a framework that uses a generative adversarial network to model kernel priors and provide better initialization, resulting in evident performance improvements over existing methods.
Blind image deconvolution (BID) is a classic yet challenging problem in the field of image processing. Recent advances in deep image prior (DIP) have motivated a series of DIP-based approaches, demonstrating remarkable success in BID. However, due to the high non-convexity of the inherent optimization process, these methods are notorious for their sensitivity to the initialized kernel. To alleviate this issue and further improve their performance, we propose a new framework for BID that better considers the prior modeling and the initialization for blur kernels, leveraging a deep generative model. The proposed approach pre-trains a generative adversarial network-based kernel generator that aptly characterizes the kernel priors and a kernel initializer that facilitates a well-informed initialization for the blur kernel through latent space encoding. With the pre-trained kernel generator and initializer, one can obtain a high-quality initialization of the blur kernel, and enable optimization within a compact latent kernel manifold. Such a framework results in an evident performance improvement over existing DIP-based BID methods. Extensive experiments on different datasets demonstrate the effectiveness of the proposed method.