CyclOT框架学习来自非配对数据的二次最优传输图

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-15 04:00

研究人员推出了一种新颖的神经网络框架CyclOT，用于从高维非配对样本中学习二次最优传输图。这种双向方法利用同步的前向-后向插值以及结合了双向二次作用、判别器约束的Jensen-Shannon端点目标和循环一致性的训练目标。该方法不需要预先计算的样本配对或显式的凸势参数化。理论结果表明，在特定条件下，该方法能够恢复最优传输图，并且在MNIST和CelebA等各种数据集上的实验验证了其性能。 AI

影响引入了一种学习最优传输图的新方法，可能改进生成模型和数据分析技术。

排序理由详细介绍一种学习最优传输图新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 13 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍一种学习最优传输图新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Shizhou Xu, Jiachen Liu, Shih-Hsin Wang, Stefan Broecker, Yuhao Huang, Bao Wang, Thomas Strohmer · 2026-09-15 04:00

CyclOT：通过同步前向-后向插值学习二次最优输运图

arXiv:2609.13892v1 Announce Type: cross Abstract: We study the recovery of forward and reverse quadratic optimal-transport maps from unpaired samples in high dimensions. We introduce a bidirectional neural framework in which the learned maps induce forward and backward displaceme…

本文如何被排名

报道来源 [1]

CyclOT：通过同步前向-后向插值学习二次最优输运图

相关实体

相关话题