visual-consistency

Tag

Cards List
#visual-consistency

TerraVis: Towards Evaluation of World-Grounded Visual Consistency in Text-to-Image Generation via MLLM Workflows

Hugging Face Daily Papers ↗ · 2026-10-02 Cached

TerraVis 提出了一个基于 MLLM 工作流的框架,用于评估文生图模型生成图像的世界一致性(如畸形结构、物理上不可能的交互、空间关系不一致等),并定义了涵盖对象、交互和场景层级的 18 类一致性违规分类体系。实验表明该框架与人类判断相关性最强,且常规指标表现优秀的模型仍可能存在大量世界一致性缺陷。

0 favorites 0 likes
#visual-consistency

StoryEngine: A State-Grounded Agentic Framework for Video Storytelling

Hugging Face Daily Papers ↗ · 2026-09-27 Cached

StoryEngine proposes a state-grounded agentic framework for video storytelling that maintains structured state representations to ensure narrative coherence and visual consistency. It demonstrates superior performance over state-of-the-art methods in comprehensive evaluations.

0 favorites 0 likes
← Back to home

Submit Feedback