图像生成 - arXivDaily 专题

1804.07739 2018-04-23 cs.CV 57%

Synthesizing Images of Humans in Unseen Poses

Guha Balakrishnan, Amy Zhao, Adrian V. Dalca, Fredo Durand, John Guttag

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments CVPR 2018

1804.00410 2018-04-03 cs.CV 57%

SyncGAN: Synchronize the Latent Space of Cross-modal Generative Adversarial Networks

Wen-Cheng Chen, Chien-Wen Chen, Min-Chun Hu

专题命中文生图：text-to-image(abstract)；分类 cs.CV

Comments 9 pages, Part of this work is accepted by IEEE International Conference on Multimedia Expo 2018

1803.11080 2018-03-30 cs.CV cs.AI cs.LG 57%

3D Consistent Biventricular Myocardial Segmentation Using Deep Learning for Mesh Generation

Qiao Zheng, Hervé Delingette, Nicolas Duchateau, Nicholas Ayache

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1612.09322 2018-03-19 cs.CV 57%

Deep Learning Logo Detection with Data Expansion by Synthesising Context

Hang Su, Xiatian Zhu, Shaogang Gong

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1802.00771 2018-02-05 cs.CV 57%

No Modes left behind: Capturing the data distribution effectively using GANs

Shashank Sharma, Vinay P. Namboodiri

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments accepted to AAAI 2018 conference

1712.07695 2017-12-22 cs.CV 57%

Adversarial Synthesis Learning Enables Segmentation Without Target Modality Ground Truth

Yuankai Huo, Zhoubing Xu, Shunxing Bao, Albert Assad, Richard G. Abramson, Bennett A. Landman

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments IEEE International Symposium on Biomedical Imaging (ISBI) 2018

1709.02073 2017-11-10 cs.CV 57%

Deep Embedding Convolutional Neural Network for Synthesizing CT Image from T1-Weighted MR Image

Lei Xiang, Qian Wang, Xiyao Jin, Dong Nie, Yu Qiao, Dinggang Shen

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1711.02245 2017-11-08 cs.CV 57%

Challenges in Disentangling Independent Factors of Variation

Attila Szabó, Qiyang Hu, Tiziano Portenier, Matthias Zwicker, Paolo Favaro

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments Submitted to ICLR 2018

1710.06270 2017-10-19 cs.CV 57%

Procedural Modeling and Physically Based Rendering for Synthetic Data Generation in Automotive Applications

Apostolia Tsirikoglou, Joel Kronander, Magnus Wrenninge, Jonas Unger

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments The project web page at http://vcl.itn.liu.se/publications/2017/TKWU17/ contains a version of the paper with high-resolution images as well as additional material

1708.00106 2017-08-10 cs.CV 57%

Material Editing Using a Physically Based Rendering Network

Guilin Liu, Duygu Ceylan, Ersin Yumer, Jimei Yang, Jyh-Ming Lien

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments 14 pages, ICCV 2017

1707.09747 2017-08-01 cs.CV 57%

Synthesis of Positron Emission Tomography (PET) Images via Multi-channel Generative Adversarial Networks (GANs)

Lei Bi, Jinman Kim, Ashnil Kumar, Dagan Feng, Michael Fulham

专题命中文生图：image synthesis(abstract)；分类 cs.CV

Comments 9 pages, 2 figures

1703.00371 2017-07-25 stat.ML cs.CR cs.MM 57%

Generating Steganographic Images via Adversarial Training

Jamie Hayes, George Danezis

专题命中文生图：image synthesis(abstract)；分类 cs.MM

Comments 9 pages

1707.07418 2017-07-25 cs.CV 57%

Generative OpenMax for Multi-Class Open Set Classification

ZongYuan Ge, Sergey Demyanov, Zetao Chen, Rahil Garnavi

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1707.04487 2017-07-25 cs.CV cs.LG 57%

Guiding InfoGAN with Semi-Supervision

Adrian Spurr, Emre Aksan, Otmar Hilliges

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1603.06078 2017-07-10 cs.GR cs.LG 57%

Deep Shading: Convolutional Neural Networks for Screen-Space Shading

Oliver Nalbach, Elena Arabadzhiyska, Dushyant Mehta, Hans-Peter Seidel, Tobias Ritschel

专题命中文生图：image synthesis(abstract)；分类 cs.GR

1612.02761 2017-04-05 cs.CV 57%

A Maximum A Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis

Yuelong Li, Chul Lee, Vishal Monga

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1511.06409 2017-01-25 cs.LG cs.CV 57%

Learning to Generate Images with Perceptual Similarity Metrics

Jake Snell, Karl Ridgeway, Renjie Liao, Brett D. Roads, Michael C. Mozer, Richard S. Zemel

专题命中文生图：image synthesis(abstract)；分类 cs.CV

1511.06078 2016-04-15 cs.CV cs.CL cs.LG 57%

Learning Deep Structure-Preserving Image-Text Embeddings

Liwei Wang, Yin Li, Svetlana Lazebnik

专题命中文生图：text-to-image(abstract)；分类 cs.CV

1505.05641 2015-05-22 cs.CV 57%

Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views

Hao Su, Charles R. Qi, Yangyan Li, Leonidas Guibas

专题命中文生图：image synthesis(abstract)；分类 cs.CV

2607.21615 2026-07-27 cs.AI cs.LG 新提交 50%

FrED: External Data Influence Estimation via Domain Knowledge Graph Grounding

FrED：通过领域知识图谱基础进行外部数据影响估计

Theodoros Aivalis, Iraklis A. Klampanos, Antonis Troumpoukis, Joemon M. Jose

机构 * National Centre for Scientific Research “Demokritos”（国家科学研究中心“德谟克利特”）； University of Glasgow（格拉斯哥大学）

专题命中文生图：image synthesis(abstract)

AI总结针对生成式AI训练数据归因问题，提出在黑盒设置下运行的概率框架，融合连续特征相似度与领域特定知识图谱，在艺术图像合成和天气预报等领域评估，证明其有效性，为外部数据影响分析提供高效可解释机制。

2607.13541 2026-07-16 cs.CR cs.LG 新提交 50%

When T2I Synthetic Data Backfires: Amplified Privacy Risks in Real-Synthetic Mix Training

当文本到图像合成数据适得其反时：真实-合成混合训练中的隐私风险放大

Na Li, Boyu Kuang, Hongsheng Hu, Liquan Chen, Hyoungshick Kim, Yansong Gao, Anmin Fu

专题命中文生图：text-to-image(abstract)

AI总结研究真实-合成混合训练中合成数据会放大真实训练样本隐私泄露问题，建立理论框架，提出RSMixLeak通过成员推理攻击评估风险，有两个变体，还给出轻量级泄露倾向指标以识别高风险数据集。

2607.10557 2026-07-14 cs.CL 新提交 50%

UNIBROWSE: A Data-to-Agent Framework for Multimodal BrowseComp

UNIBROWSE：用于多模态浏览比较的数据到智能体框架

Xiyu Wei, Qingwei Zong, Zhuocheng Yu, Sujian Li

机构 * Key Laboratory of Computational Linguistics, MOE, Peking University（教育部计算语言学重点实验室，北京大学）； School of Software and Microelectronics, Peking University（北京大学软件与微电子学院）； School of Computer Science, Peking University（北京大学计算机科学学院）

专题命中文生图：text-to-image(abstract)

AI总结研究多模态浏览比较任务，提出UNIBROWSE统一数据管道，同时生成三种模式训练数据，增强知识图谱，引入探索度度量。经此训练的35B规模智能体在多模态浏览比较基准测试中性能领先，超多个闭源智能体工作流程。

Comments 17 pages, 5 figures

2607.04339 2026-07-07 cs.LG cs.AI cs.CR 新提交 50%

One Framework for All: Cross-Modal Membership Inference for Generative Models

一框架适用于所有：生成模型的跨模态成员推理

Dayong Ye, Tainqing Zhu, Kun Gao, Junhao Liu, Yichuan Chen, Shuai Zhou, Hengzhu Liu, Bo Liu, Wanlei Zhou

专题命中文生图：text-to-image(abstract)

AI总结研究生成模型的跨模态成员推理问题，基于生成模型输出分布可近似训练数据分布的特性，在共享嵌入空间建模，通过似然比测试推理，贡献统一框架，性能优于现有方法。

2606.31711 2026-07-01 cs.AI 新提交 50%

Arena-T2I Hard: Benchmarking and Improving Faithfulness with Dependency-Aware Checklist

Arena-T2I Hard：基于依赖感知清单的忠实性基准测试与改进

Yuanhao Ban, Tong Xie, Sohyun An, Yunqi Hong, Evan Frick, I-Hung Hsu, Wei-Lin Chiang, Ion Stoica, Cho-Jui Hsieh

机构 * Arena Intelligence Inc ； UCLA（加州大学洛杉矶分校）

专题命中文生图：text-to-image(abstract)

AI总结针对现有忠实性基准在复杂指令上的不足，提出310条压力测试提示集Arena-T2I Hard，并设计依赖感知清单奖励结合美学奖励，显著提升文本到图像模型的忠实性-美学权衡。

2606.28221 2026-06-29 astro-ph.IM 新提交 50%

Faraday Tomography with the SKA: A New Era of Cosmic Magnetism Studies

SKA的法拉第层析成像：宇宙磁学研究的新时代

Miguel Carcamo, Anna M. M. Scaife, Jeroen Stil, Russ Taylor, Jennifer L. West, Tessa Vernstrom

专题命中文生图：image synthesis(abstract)

AI总结本文综述了SKA通过法拉第旋转和法拉第测量合成技术研究宇宙磁场的能力，重点介绍了AA4阶段的配置及其在星系、星系团和宇宙网中高分辨率法拉第层析成像的应用。

Comments Published in Advancing Astrophysics with the SKA II (AASKAII), 2026 (arXiv:2606.20366). Report-no:AASKAII/Carcamo01. Advancing Astrophysics with the SKA II (AASKAII) outlines the transformative scientific advances that will be enabled by the SKA telescopes

2606.23645 2026-06-23 math-ph math.MP 新提交 50%

Which Waveguide Network Realizes a Prescribed Transmission Profile? An Exact Forward Construction

哪个波导网络实现指定的传输轮廓？一种精确的正向构造方法

Tristan M. Lawrie

专题命中文生图：image synthesis(abstract)

AI总结提出一种基于规范平移亥姆霍兹算子的周期性波导网络散射特性的可解析反演框架，通过傅里叶展开将晶格电抗映射到图架构，直接从目标传输系数确定物理结构，实现从一维角滤波到二维图像合成的精确波前构造。

Comments 12 pages, 4 figures

2606.12972 2026-06-12 cs.HC 新提交 50%

From Prompts to Preferences: An Open-Source Platform for Generative AI-Enhanced Conjoint Analysis

从提示到偏好：生成式AI增强联合分析的开源平台

Philipp Brauner

专题命中文生图：text-to-image(abstract)

AI总结提出一个开源、自托管的联合分析调查平台，利用生成式AI（大语言模型和文本到图像模型）创建集成刺激格式，降低研究门槛，并通过概念验证研究展示其有效性。

2606.03354 2026-06-03 cs.CR 50%

ImageAuditor: Membership Inference Attack against Image-based Retrieval-Augmented Generation

ImageAuditor: 针对基于图像的检索增强生成系统的成员推理攻击

Jinghuai Zhang, Pengyue Yu, Zhexiao Lin, Kunlin Cai, Fnu Suya, Yuan Tian

专题命中文生图：text-to-image(abstract)

AI总结提出首个针对基于图像的检索增强生成（IRAG）的成员推理攻击方法ImageAuditor，通过奖励引导策略优化解决跨模态检索和判别信号提取两大挑战，在仅需四次查询时AUROC超过80%。

2501.12374 2026-05-29 cs.HC cs.AI cs.CY 50%

Expertise elevates AI usage: experimental evidence comparing laypeople and professional artists

专业知识提升AI使用：比较普通人和专业艺术家的实验证据

Thomas F. Eisenmann, Andres Karjus, Mar Canet Sola, Levin Brinkmann, Bramantyo Ibrahim Supriyatno, Iyad Rahwan

机构 * Center for Humans and Machines, Max Planck Institute for Human Development（人类与机器中心，马克斯·普朗克人类发展研究所）； Tallinn University（塔林大学）； Estonian Business School（爱沙尼亚商学院）； Academy of Media Arts Cologne（科隆媒体艺术学院）

专题命中文生图：text-to-image(abstract)

AI总结通过实验比较50位专业艺术家和普通人使用生成式AI进行图像复制和创意生成的表现，发现艺术家的专业技能迁移到AI使用中，在复制准确性和发散思维上均优于普通人，而GPT-4o在创意任务上平均略优于艺术家但未超越最佳人类。

Comments Eisenmann and Karjus contributed equally to this work and share first authorship

Journal ref International Journal of Human-Computer Interaction, 2026, pp 1-22

2604.23703 2026-05-26 cs.HC cs.AI cs.CY 50%

Talking Slide Avatars: Open-Source Multimodal Communication Approach for Teaching

Talking Slide Avatars: 面向教学的开源多模态通信方法

Xinxing Wu

机构 * School of Mathematics and Computer Science, Kentucky State University（肯塔基州立大学数学与计算机科学学院）

专题命中文生图：image synthesis(abstract)

AI总结提出一种集成OpenVoice和Ditto-TalkingHead的开源工作流，用于创建可说话的幻灯片头像，以增强在线教学中的教师存在感和叙事连续性。

Comments 15 pages

1. 文生图 3480 篇

Synthesizing Images of Humans in Unseen Poses

SyncGAN: Synchronize the Latent Space of Cross-modal Generative Adversarial Networks

3D Consistent Biventricular Myocardial Segmentation Using Deep Learning for Mesh Generation

Deep Learning Logo Detection with Data Expansion by Synthesising Context

No Modes left behind: Capturing the data distribution effectively using GANs

Adversarial Synthesis Learning Enables Segmentation Without Target Modality Ground Truth

Deep Embedding Convolutional Neural Network for Synthesizing CT Image from T1-Weighted MR Image

Challenges in Disentangling Independent Factors of Variation

Procedural Modeling and Physically Based Rendering for Synthetic Data Generation in Automotive Applications

Material Editing Using a Physically Based Rendering Network

Synthesis of Positron Emission Tomography (PET) Images via Multi-channel Generative Adversarial Networks (GANs)

Generating Steganographic Images via Adversarial Training

Generative OpenMax for Multi-Class Open Set Classification

Guiding InfoGAN with Semi-Supervision

Deep Shading: Convolutional Neural Networks for Screen-Space Shading

A Maximum A Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis

Learning to Generate Images with Perceptual Similarity Metrics

Learning Deep Structure-Preserving Image-Text Embeddings

Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views

FrED: External Data Influence Estimation via Domain Knowledge Graph Grounding

When T2I Synthetic Data Backfires: Amplified Privacy Risks in Real-Synthetic Mix Training

UNIBROWSE: A Data-to-Agent Framework for Multimodal BrowseComp

One Framework for All: Cross-Modal Membership Inference for Generative Models

Arena-T2I Hard: Benchmarking and Improving Faithfulness with Dependency-Aware Checklist

Faraday Tomography with the SKA: A New Era of Cosmic Magnetism Studies

Which Waveguide Network Realizes a Prescribed Transmission Profile? An Exact Forward Construction

From Prompts to Preferences: An Open-Source Platform for Generative AI-Enhanced Conjoint Analysis

ImageAuditor: Membership Inference Attack against Image-based Retrieval-Augmented Generation

Expertise elevates AI usage: experimental evidence comparing laypeople and professional artists

Talking Slide Avatars: Open-Source Multimodal Communication Approach for Teaching