{"ID":2896836,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.05604","arxiv_id":"2507.05604","title":"Kernel Density Steering: Inference-Time Scaling via Mode Seeking for Image Restoration","abstract":"Diffusion models show promise for image restoration, but existing methods often struggle with inconsistent fidelity and undesirable artifacts. To address this, we introduce Kernel Density Steering (KDS), a novel inference-time framework promoting robust, high-fidelity outputs through explicit local mode-seeking. KDS employs an $N$-particle ensemble of diffusion samples, computing patch-wise kernel density estimation gradients from their collective outputs. These gradients steer patches in each particle towards shared, higher-density regions identified within the ensemble. This collective local mode-seeking mechanism, acting as \"collective wisdom\", steers samples away from spurious modes prone to artifacts, arising from independent sampling or model imperfections, and towards more robust, high-fidelity structures. This allows us to obtain better quality samples at the expense of higher compute by simultaneously sampling multiple particles. As a plug-and-play framework, KDS requires no retraining or external verifiers, seamlessly integrating with various diffusion samplers. Extensive numerical validations demonstrate KDS substantially improves both quantitative and qualitative performance on challenging real-world super-resolution and image inpainting tasks.","short_abstract":"Diffusion models show promise for image restoration, but existing methods often struggle with inconsistent fidelity and undesirable artifacts. To address this, we introduce Kernel Density Steering (KDS), a novel inference-time framework promoting robust, high-fidelity outputs through explicit local mode-seeking. KDS em...","url_abs":"https://arxiv.org/abs/2507.05604","url_pdf":"https://arxiv.org/pdf/2507.05604v2","authors":"[\"Yuyang Hu\",\"Kangfu Mei\",\"Mojtaba Sahraee-Ardakan\",\"Ulugbek S. Kamilov\",\"Peyman Milanfar\",\"Mauricio Delbracio\"]","published":"2025-07-08T02:33:44Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"eess.IV\"]","methods":"[\"Diffusion Model\"]","has_code":false}