Hongjie Wu 邬鸿杰

Craniofacial reconstruction predicts craniofacial features from skull morphology to reconstruct the craniofacial surface. Although previous methods have achieved promising performance, they face three critical limitations: insufficient image quality, poor identity preservation, and difficulties with conditional control. To overcome these challenges, we propose a novel diffusion-based craniofacial reconstruction method that preserves identity across domain transfer. Our approach incorporates multiple modules to separately manage multimodal information including skull data, landmarks, texture features, and biometric information, yielding high-fidelity results under various constraints. Furthermore, by enabling flexible modification of biometric information through standardized text prompts, our method achieves fine-grained control while maintaining individual identity characteristics. Extensive experimental results demonstrate that our method outperforms existing approaches in image quality and identity retrieval, showcasing exceptional robustness, strong identity preservation, and enhanced editability. Our code is available at: https://github.com/mqzhang2024/ICCR.

Diffusion models have shown remarkable promise for image restoration by leveraging powerful priors. Prominent methods typically frame the restoration problem within a Bayesian inference framework, which iteratively combines a denoising step with a likelihood guidance step. However, the interactions between these two components in the generation process remain underexplored. In this paper, we analyze the underlying gradient dynamics of these components and identify significant instabilities. Specifically, we demonstrate conflicts between the prior and likelihood gradient directions, alongside temporal fluctuations in the likelihood gradient itself. We show that these instabilities disrupt the generative process and compromise restoration performance. To address these issues, we propose Stabilized Progressive Gradient Diffusion (SPGD), a novel gradient management technique. SPGD integrates two synergistic components: (1) a progressive likelihood warm-up strategy to mitigate gradient conflicts; and (2) adaptive directional momentum (ADM) smoothing to reduce fluctuations in the likelihood gradient. Extensive experiments across diverse restoration tasks demonstrate that SPGD significantly enhances generation stability, leading to state-of-the-art performance in quantitative metrics and visually superior results. Code is available at \hrefhttps://github.com/74587887/SPGDhere.

Diffusion models have demonstrated remarkable efficacy in generating high-quality samples. Existing diffusion-based image restoration algorithms exploit pre-trained diffusion models to leverage data priors, yet they still preserve elements inherited from the unconditional generation paradigm. These strategies initiate the denoising process with pure white noise and incorporate random noise at each generative step, leading to over-smoothed results. In this paper, we introduce a refined paradigm for diffusion-based image restoration. Specifically, we opt for a sample consistent with the measurement identity at each generative step, exploiting the sampling selection as an avenue for output stability and enhancement. Besides, we start the restoration process with an initialization combined with the measurement signal, providing supplementary information to better align the generative process. Extensive experimental results and analyses validate the effectiveness of our proposed approach across diverse image restoration tasks.