A terminal-side media cover generation method based on multi-modal metadata driving

By using a multimodal metadata-driven terminal-side generation method, combining terminal interface layout and text semantic features, a diffusion model initialization parameter is constructed to generate and verify the cover image. This solves the adaptation and compliance issues of cover generation in existing technologies and achieves efficient and secure terminal-side cover image generation.

CN122415784APending Publication Date: 2026-07-17LERONG SMART HOME (BEIJING) TECHNOLOGY DEVELOPMENT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
LERONG SMART HOME (BEIJING) TECHNOLOGY DEVELOPMENT CO LTD
Filing Date
2026-03-31
Publication Date
2026-07-17

Smart Images

  • Figure CN122415784A_ABST
    Figure CN122415784A_ABST
Patent Text Reader

Abstract

本公开的实施例提供了一种基于多模态元数据驱动的终端侧媒体封面生成方法，应用于人工智能与多媒体处理技术领域。该方法包括接收服务端下发的多模态元数据包，在终端侧对其进行解析与反量化处理以还原视觉特征，并基于终端界面布局与元数据中的适配基准参数计算视觉安全区；依据元数据中的文本特征生成提示词，结合视觉特征转换所得的偏置参数与视觉安全区构建包含构图约束的模型初始化参数；将所述初始化参数输入本地生成模型进行图像推理，并对生成的封面图像进行安全合规校验，校验通过后予以展示。以此方式，可以解决传统的媒体封面生成技术中网络依赖性强、无法适配终端多样化的界面布局、且难以兼顾个性化与内容合规性的问题。
Need to check novelty before this filing date? Find Prior Art