Digital human real-time interaction method based on large model and RAG

CN119990309APending Publication Date: 2025-05-13SUZHOU ZHONGDIHANG INFORMATION TECH CO LTD

Patent Information

Application Number
CN202510013257.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-06
Publication Date
2025-05-13

Smart Images

  • Figure CN119990309A_ABST
    Figure CN119990309A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of digital human interaction, and discloses a digital human real-time interaction method based on a large model and an RAG, voice semantics input by a user are analyzed based on an ASR model, an analysis result is input into an RAG module, a data segment related to user instruction semantics is retrieved from a knowledge base through the RAG module, and the data segment is stored in the RAG module. And the answer text is generated through the large model, so that the accuracy of the answer text and the correlation of the context are improved. Besides, the knowledge base also supports business scenarized configuration, and in the reasoning process of a large model, reasoning of an answer text is carried out based on a multi-modal vector including voice features and gesture information input by a user, so that higher requirements on semantic comprehension and multi-modal interaction in a complex business scene are met.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Large model voice question-answering system for multi-source heterogeneous local knowledge base

    CN118193694A

  • Human-computer interaction method, device and equipment based on digital human, storage medium and product

    CN118377868A

  • Digital human interaction method and device in virtual space, equipment, medium and product

    CN118445458A

  • Artificial intelligence-driven multifunctional English language learning and evaluation method and application thereof

    CN118535683A

  • Text extraction method and system based on large language model

    CN118964575A

Cited By

  • Controllable digital human real-time interaction system based on multi-modal sensor

    CN120631181A

  • End-side collaborative lightweight voice interaction large model optimization method and system

    CN121148381A

  • Interactive digital human multi-modal display method based on large model RAG and pre-rendering matching

    CN121479013A

  • Multi-modal display method of interactive digital human based on large model RAG and pre-rendering matching

    CN121479013B

  • Financial customer service RAG knowledge base updating method and device and storage medium

    CN121882192A