面向大语言模型的信息检索新视角，首次提出以去噪为核心，提升检索质量与鲁棒性
arXiv:2605.00505v2 Announce Type: replace-cross Abstract: Modern information retrieval (IR) is no longer consumed primarily by humans but increasingly…
用深度学习去噪技术提升AI心电图分析精度，论文解读最新方法
arXiv:2605.03183v2 Announce Type: replace Abstract: Evaluating canine electrocardiograms (ECGs) is challenging due to noise that can obscure clinicall…
提出归一化等变性的结构先验，可应用于任意骨干网络的图像去噪，有效提升分布偏移健壮性。
arXiv:2605.08193v2 Announce Type: replace-cross Abstract: Normalization Equivariance (NE) is a structural prior that improves robustness to distributi…
用神经网络实时去除头发渲染中的噪点，大幅提升动画与游戏画质。
arXiv:2605.17557v1 Announce Type: cross Abstract: We propose a lightweight real-time method for reconstructing strand-based hair G-Buffers from severe…
提出自适应双阶段ROF去噪模型，突破经典方法对非均匀噪声的局限，理论新颖性强
arXiv:2510.04382v2 Announce Type: replace-cross Abstract: Even though more than 30 years have passed since the seminal Rudin--Osher--Fatemi (ROF) pape…
动态分块技术让扩散语言模型根据内容自动划分块，取代固定位置分块，增强语义连贯性，提升生成效率。
arXiv:2605.15676v1 Announce Type: new Abstract: Block discrete diffusion language models factorize a sequence autoregressively over fixed-size positio…
一键直达arXiv最新论文，探讨扩散模型、去噪架构与创意
arXiv:2605.16415v1 Announce Type: cross Abstract: The creativity of diffusion models refers to their ability to generate highly realistic images that …
提出残差噪声学习方法，解决PET去噪中跨剂量泛化难题，有效缓解平均效应带来的退化。
arXiv:2604.16925v3 Announce Type: replace Abstract: Cross-dose denoising for low-dose positron emission tomography (LDPET) has been proposed to addres…