一种语音识别的方法和装置
By generating error-prone phrase pairs to optimize the speech recognition model, the problems of high cost and low efficiency in existing technologies are solved, the accuracy and efficiency of speech recognition are improved, and the effect of human-computer language interaction system is enhanced.
CN116469389BActive Publication Date: 2026-07-17JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
- Filing Date
- 2023-03-28
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
The optimization of existing speech recognition models is costly and inefficient, with low accuracy and efficiency, which affects the effectiveness of human-computer language interaction systems and reduces user experience.
Method used
By mining abnormal data to generate error-prone phrase pairs, the speech recognition model can be optimized. The accuracy and efficiency of the speech recognition model can be improved by using error-prone phrase pairs to optimize the speech recognition model.
Benefits of technology
It reduced the cost of optimizing the model, improved the accuracy and efficiency of the speech recognition model, optimized the effect of the human-computer language interaction system, and enhanced the user experience.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116469389B_ABST
Abstract
本发明公开了一种语音识别的方法和装置，涉及人工智能技术领域。该方法的一具体实施方式包括：根据第一文本数据对应的多个语音数据，通过语音识别模型生成多个第二文本数据；对于每一第二文本数据，在第二文本数据与第一文本数据不一致的情况下，通过第一文本数据和第二文本数据生成短语数据对；对所有短语数据对进行数据挖掘处理，生成易错短语对，并利用易错短语对优化语音识别模型；使用优化后的语音识别模型进行语音识别。该实施方式能够通过挖掘异常数据得到高质量的训练数据，降低优化模型所耗费的成本，提高优化模型的效率，并且在使用时可以提高语音识别模型的准确率和效率，从而优化人机语言交互系统的效果，提高用户的使用体验。
Need to check novelty before this filing date? Find Prior Art