An intent deconstruction-based cross-modal content dynamic generation method and system

By employing intent deconstruction and dynamic programming methods, the controllability and coherence issues in generating long videos in existing technologies are resolved, enabling efficient content generation through human-computer collaboration, and making it applicable to various professional fields.

CN122293947APending Publication Date: 2026-06-26SHENZHEN HEJIUGUIYI CULTURAL DEVELOPMENT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-19
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing end-to-end generation models lack controllability, coherence, and process control when generating long videos, and have limited generation capabilities, making it difficult to meet the logical consistency of complex content and the requirements for long videos.

Method used

The intent deconstruction engine transforms user input into a structured intent blueprint, a dynamic programming engine plans the generation pipeline, and the pipeline execution engine executes it, realizing a multi-step content generation process, including rapid generation of short videos and modular generation of long videos.

Benefits of technology

It achieves fine-grained and precise control over content, ensures the internal coherence of complex content, supports human-computer collaborative creation processes, and can generate long videos, making it suitable for content generation in various professional fields.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122293947A_ABST
    Figure CN122293947A_ABST
Patent Text Reader

Abstract

This invention discloses a method and system for dynamic cross-modal content generation based on intent deconstruction. This method deconstructs the user's creative intent, which may be insufficiently defined due to brief descriptions, into structured intent components that can be precisely executed by the machine. Then, based on these components, a customized content generation pipeline is dynamically planned, and finally, the pipeline is executed to generate the final content. By introducing a "deconstruct-plan-generate" paradigm, this invention transforms content creation from an uncontrollable, probabilistic process into a structured, predictable, interventionist, and collaborative industrial process, addressing the fundamental shortcomings of existing technologies when handling complex, logically strong, and highly consistent content. This invention can quickly generate short videos and, through a modular "building block" approach, can also generate cinematic-quality videos exceeding 15 minutes in length, with complete narratives and a consistent style.
Need to check novelty before this filing date? Find Prior Art