AI Robot Speech Intent Learning via User Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional artificial intelligence systems require error correction by managers to accurately analyze user intent from speech recognition, limiting their ability to learn and improve independently.
Innovation Solution
An AI-based method that receives user speech information, processes it through a pre-learned intent analysis model, and generates a second intent analysis model by incorporating user evaluation feedback, allowing the system to learn and improve without manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional speech recognition systems use pre-learned intent analysis models, then speech processing speed is improved, but accuracy in determining user intent deteriorates when errors occur without manual correction
Solution Approach 1:
The patent implements a feedback mechanism where the system receives correction information from users when intent determination is inaccurate. This correction information is stored and used to retrain the intent analysis model, creating a continuous improvement loop that maintains both fast processing and high accuracy over time
Solution Approach 2:
The system performs self-learning by automatically incorporating user corrections into model retraining without requiring external manual intervention. The robot independently improves its intent analysis capabilities by processing its own performance data and applying learned improvements to future speech recognition tasks
2Measurement precision
If manual error correction by managers is implemented, then intent analysis accuracy is improved, but system complexity and operational burden increase
Solution Approach 1:
The system eliminates the need for manual manager intervention by implementing self-correction capabilities. Users directly provide corrections that are automatically incorporated into model retraining, transforming a complex manual process into a simple user feedback mechanism
Solution Approach 2:
The patent establishes an automated feedback loop where correction information flows directly from users to the learning system, bypassing manual management layers. This reduces operational complexity while maintaining accuracy improvement
3Measurement precision
If deep learning models are retrained with user feedback, then learning accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The system applies partial retraining by using only the accumulated correction information rather than complete dataset retraining. This selective approach improves accuracy on specific error patterns while minimizing overall retraining time and computational overhead
Solution Approach 2:
The system accumulates correction information over time before initiating retraining operations. This preliminary data collection allows for more efficient batch processing and reduces the frequency and duration of full retraining cycles
Data Source
AI summary
Disclosed is a speech recognition method of an artificial intelligence robot. The speech recognition method includes: receiving uttered speech information of a user from an external device; inputting the speech information to a pre-learned first intent analysis model, and determining an utterance intent of the user according to an output value of the first intent analysis; transmitting response information corresponding to the determined utterance intent of the user to the external device; receive evaluation information of the user on the response information from the external device; and generating a second intent analysis model by adding the evaluation information to learning data and learning the first intent analysis model. Accordingly, an intelligent device is capable of learning an accurate utterance intent even without error correction by a manager.