Voice Trigger Detection Using Dynamic Dictionary Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional speech recognition systems face challenges in detecting triggers and performing speech recognition in continuous speech, often failing to recognize voice commands accurately after the initial trigger is detected.
Innovation Solution
An information processing apparatus that includes a controller to detect triggers from voice signals and perform speech recognition using a trigger and speech recognition dictionary, allowing for continuous recognition of voice commands by switching between trigger and speech recognition dictionaries based on detected triggers and silence sections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional speech recognition systems use a single speech recognition dictionary, then the system structure is simple, but the system fails to detect triggers and perform speech recognition accurately in continuous speech
Solution Approach 1:
The patent divides the speech recognition process into distinct phases by using multiple dictionaries: a trigger dictionary for detecting trigger words and a speech recognition dictionary for recognizing commands. This segmentation allows the system to handle different functional requirements separately, improving accuracy in continuous speech while maintaining manageable complexity through structured organization.
Solution Approach 2:
The patent implements dynamic switching between different dictionaries based on the speech recognition state. The controller selectively switches between trigger dictionary and speech recognition dictionary according to whether trigger detection is needed or command recognition is needed, enabling the system to adapt to continuous speech patterns dynamically.
2Adaptability or versatility
If the system switches between trigger and speech recognition dictionaries, then continuous speech recognition is enabled, but the system complexity increases
Solution Approach 1:
The controller dynamically switches between trigger dictionary and speech recognition dictionary based on the current speech recognition state. When trigger detection is needed, the trigger dictionary is selected; when command recognition is needed, the speech recognition dictionary is selected. This dynamic adaptation enables continuous speech recognition while managing complexity through state-based control.
Solution Approach 2:
The system uses feedback from the speech recognition process to determine which dictionary to use. The controller monitors the speech recognition state and adjusts dictionary selection accordingly, creating a feedback loop that enables continuous speech recognition while keeping the control logic manageable through state-aware decision making.
Data Source
AI summary
According to one embodiment, an information processing apparatus includes one or more processors configured to detect a trigger from a voice signal, the trigger indicating start of voice recognition; and to perform voice recognition of a recognition sound section subsequent to a trigger sound section including the detected trigger, referring to a trigger and voice recognition dictionary corresponding to the trigger.