基于用户反馈的AIGC生成内容自适应方法及系统
By combining neural Hawkes processes and deep kernel learning with Bayesian optimization algorithms, the generative model is dynamically adjusted to adapt to multimodal user feedback, which solves the shortcomings of existing systems in adaptive optimization and improves the accuracy and adaptability of generated content.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN QIANBO TECH CO LTD
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing generative AI systems lack the ability to understand and adapt to complex user intentions in terms of adaptive optimization of multimodal user feedback, resulting in insufficient accuracy of generated content.
Event modeling is performed using neural Hawkes processes, combined with deep kernel learning and Bayesian optimization algorithms. By using causal influence datasets and counterfactual reasoning, the generative model is dynamically adjusted to adapt to user feedback, generating optimized AIGC content.
It enables explicit characterization of the temporal dynamics and potential causal dependencies of user interactions, enhances the understanding of complex user intentions, and improves the accuracy and adaptability of generated content.
Smart Images
Figure CN122114191B_ABST