paper-with-me

홈 › Papers

A Principled Self-Referenced Early Stopping Approach for Deep Image Prior

2026-05-24 · Chaoyan Huang, Cheng-Han Huang, Ismail R. Alkhouri, Rongrong Wang arxiv

Recently, Deep Image Prior (DIP) has demonstrated strong capabilities for solving inverse imaging problems (IIPs) by optimizing a randomly initialized convolutional neural network in a training-data-free regime. However, DIP suffers from overfitting to noisy measurements due to network over-parameterization, making early stopping (ES) essential. The most successful ES method tracks fluctuations in the running variance of the network output to detect overfitting. However, in many applications, these fluctuations may appear prematurely, leading to unstable reconstructions. In this paper, we first show that nearly optimal DIP early stopping can be achieved when two independent noisy copies of the degraded image are available. Motivated by this observation, and since obtaining two fully independent copies is infeasible, we propose an overfitting detection framework based on constructing pseudo self-referenced images, resulting in three IIP-specific algorithms. Our approach is further supported by theoretical results on single-reference validation, pseudo-validation estimation, and the impact of shared noise. Across different IIPs, ranging from natural image restoration to medical image reconstruction, and under varying noise levels and noise types, our methods consistently outperform existing DIP early stopping approaches, all without requiring an accurate estimate of the noise level.

📄 PDF Abstract BibTeX arXiv:2605.25299

Code (0)

등록된 구현이 없습니다.

Tasks

Image ReconstructionImage Restoration

Similar Papers 제목 키워드 기반

Self-Validation: Early Stopping for Single-Instance Deep Generative Priors

2021-10-23 · Taihui Li, Zhong Zhuang, Hengyue Liang, Le Peng 외

Recent works have shown the surprising effectiveness of deep generative models in solving numerous image reconstruction (IR) tasks, even without training data. We call these models, such as deep image prior and deep deco…

DecoderImage Reconstruction

Statistical Early Stopping for Reasoning Models

2026-02-15 · Yangxinyu Xie, Tao Wang, Soham Mallick, Yan Sun 외 arxiv

While LLMs have seen substantial improvement in reasoning capabilities, they also sometimes overthink, generating unnecessary reasoning steps, particularly under uncertainty, given ill-posed or ambiguous queries. We intr…

Optimal Stopping in Latent Diffusion Models

2025-10-09 · Yu-Han Wu, Quentin Berthet, Gérard Biau, Claire Boyer 외 arxiv

We identify and analyze a surprising phenomenon of Latent Diffusion Models (LDMs) where the final steps of the diffusion can degrade sample quality. In contrast to conventional arguments that justify early stopping for n…

Dimensionality Reduction

Rethinking Deep Image Prior for Denoising

2021-08-29 · ICCV 2021 10 · Yeonsik Jo, Se Young Chun, Jonghyun Choi

Deep image prior (DIP) serves as a good inductive bias for diverse inverse problems. Among them, denoising is known to be particularly challenging for the DIP due to noise fitting with the requirement of an early stoppin…

DenoisingInductive BiasSSIM

Bigger Isn't Always Memorizing: Early Stopping Overparameterized Diffusion Models

2025-05-22 · Alessandro Favero, Antonio Sclocchi, Matthieu Wyart

Diffusion probabilistic models have become a cornerstone of modern generative AI, yet the mechanisms underlying their generalization remain poorly understood. In fact, if these models were perfectly minimizing their trai…

Memorization