Speech interaction method, speech interaction system and storage medium
By combining interactive voice and text, and using multiple emotion feature extraction models and multilayer perceptrons to generate response voices with whole sentences and local prosodic features, the problem of intelligent dialogue systems being unable to recognize user emotions is solved, and emotionally rich and natural voice interaction is achieved.
JP2026027170AActive Publication Date: 2026-02-18NANJING SILICON INTELLIGENCE TECH CO LTD
4 Cites 1 Cited by
Patent Information
- Application Number
- JP2025006547
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-08-05
- Filing Date
- 2025-01-17
- Publication Date
- 2026-02-18
- Estimated Expiration
- 2045-01-17
AI Technical Summary
Technical Problem
Existing intelligent dialogue systems are unable to effectively recognize users' emotions and control the emotional changes in their response voices, resulting in insufficient emotional communication and overly simplistic responses.
Method used
By combining interactive voice and interactive text, using multiple emotion feature extraction models and multilayer perceptrons for emotion classification, responsive voice with whole sentence and local prosodic features is generated, achieving emotionally rich and natural voice interaction.
Benefits of technology
It improves the accuracy of emotion classification and the naturalness of generated response speech, enhancing the user's emotional communication experience.
✦ Generated by Eureka AI based on patent content.
Abstract
To provide a voice interaction method, a voice interaction system and a storage medium capable of improving richness and naturalness of feeling.SOLUTION: Receiving an interaction speech input by a user, determining an emotion tag corresponding to the interaction speech according to the interaction speech and an interaction text corresponding to the interaction speech, determining a response text corresponding to the interaction text and a first prosodic feature and a second prosodic feature corresponding to the response text according to the emotion tag, and generating and outputting a response speech corresponding to the interaction speech according to the response text, the first prosodic feature, and the second prosodic feature; The first prosodic feature represents a prosodic feature of an entire sentence of the response text, and the second prosodic feature represents a local prosodic feature of each character in the response text.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art