English(EN) KINESYS: Test-Time Adaptation for Reference-Based Effect Editing in Images

KINESYS框架通过在效果迁移过程中保留源内容来增强图像编辑

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-27 04:00

研究人员开发了KINESYS，一个新颖的用于效果参考编辑的测试时自适应框架。该系统旨在在保留源图像的原始内容和构图的同时，准确地从参考图像迁移所需视觉效果。KINESYS采用强大的优化策略，通过在多个噪声样本和优化运行中微调生成模型，以实现多样化和忠实的编辑。评估表明，KINESYS在源内容保留方面优于当前最先进的图像编辑器，并在效果迁移方面与Nano Banana 2等模型具有竞争力。 AI

影响该框架可以提高AI驱动的图像编辑工具的保真度和控制力，从而实现更精确的视觉效果应用。

排序理由该集群描述了一篇关于图像编辑新颖框架的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CV 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: 8 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了一篇关于图像编辑新颖框架的最新研究论文。[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, product; 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: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CV TIER_1 English(EN) · Noah Frahm, Luchao Qi, Haoze Li, Dan B Goldman, Sambit Bhattacharya, Roni Sengupta · 2026-10-27 04:00

KINESYS：基于参考的图像效果编辑的测试时自适应

arXiv:2610.33292v1 Announce Type: new Abstract: Visual effects are often easier to demonstrate via reference than to describe in words. We study effect-reference editing, where a reference image shows a desired visual effect and language identifies which properties should guide t…

本文如何被排名

报道来源 [1]

KINESYS：基于参考的图像效果编辑的测试时自适应

相关实体

相关话题