多模态内容自适应生成方法、终端设备以及存储介质
By using a multimodal content adaptive generation method, the problem of cross-platform content style transfer was solved, enabling efficient generation of adapted content on different e-commerce platforms and improving content dissemination efficiency and reuse rate.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN BIT EMERGENCE TECHNOLOGY CO LTD
- Filing Date
- 2025-08-21
- Publication Date
- 2026-07-17
AI Technical Summary
The lack of an effective cross-platform content style transfer mechanism in existing technologies means that content needs to be created separately for different e-commerce platforms, which consumes a lot of time and energy and has a very low content reuse rate, thus reducing the efficiency of content dissemination.
By using a multimodal content adaptive generation method, the feature vectors of the target platform and user feature vectors are determined, and a multimodal generation model is used to generate content adapted to different platforms, including data collection, analysis, user behavior data processing, and application of the multimodal generation model.
It achieves precise cross-platform content adaptation, improves content dissemination efficiency, reduces time and effort consumption, and increases content reuse rate.
Smart Images
Figure CN121030091B_ABST