Speech Recognition TRIZ Cases
- Mobile Terminal Search Interface with Voice Input and Engine Selection
A mobile terminal search interface integrates voice recognition and virtual regions for selecting search engines.
- Digital Assistant Call Sign Detection for Unanswered Police Radio Calls
A digital assistant identifies wake words as call signs to manage device operations when users do not answer calls.
- Environmental Detector Validating Personalized Voice Alerts
A detector emits personalized verbal alerts in a user's recorded voice to capture occupant attention during emergencies.
- Transcription System Audio-Text Position Correlation
A transcription system correlates text positions with audio segments using selectable interface elements for direct playback navigation.
- Electronic Apparatus Sensor Prioritization for Speech Recognition
An electronic apparatus identifies an effective sensor device based on audio signal similarity to perform speech recognition.
- Automatic Level Control for Speech Signals Using Sub-Frame Gain
Sub-frame analysis distinguishes speech from noise to adjust amplification, resolving inconsistent audio quality without excessive computational complexity.
- Context-Aware Voice Assistant Meeting Prompts
A voice assistant analyzes speech parameters to deliver proactive notifications during meetings.
- Voice Extraction via Microphone Array Feature Fusion
Fusing normalized beamforming features with blind source separation outputs reduces environmental noise and improves speech extraction accuracy.
- Voice Input Apparatus with Display Feedback for Error Correction
A voice input apparatus displays recognized processing information on a screen to allow operator verification and modification before execution.
- Voice Interface Dynamic Endpoint Switching
Dynamic endpoint selection eliminates central control bottlenecks while maintaining high voice quality.
- Voice Command Visual Confirmation GUI Feedback
A speech recognition graphical user interface changes element formats to visually confirm recognized voice commands.
- RNN-T Encoder Customization via Audio Synthesis
Synthesizing domain audio updates RNN-T encoders to adapt vocabularies without pre-computed alignments.
- Multimodal Information Processing Device Adaptability
An abstract output information generator separates processing logic from hardware specifics, eliminating redesign needs when output devices change.
- Speech Pitch Estimation Using Mixture Gaussian Distributions
A pitch estimating apparatus computes normalized autocorrelation and generates mixture Gaussian distributions for candidate pitches.
- Voice Command Response Control Using Relative Receipt Time
System uses relative voice command receipt time to prevent unintended actions in multi-device environments by triggering only the first recipient.