Audible Tap Detection via Linear Prediction Error Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user interfaces struggle to reliably detect and respond to non-verbal user interactions such as audible taps, double-taps, and other sound sequences, which are essential for natural user actions in augmented reality environments.
Innovation Solution
The system employs linear prediction coding to isolate audible taps from microphone audio signals, analyzing prediction error signals for dynamic range and periodicity to detect and validate these events, and uses multiple microphones for sound localization to differentiate user-generated sounds from noise or echoes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional speech recognition is used to detect user commands, then verbal commands can be recognized, but non-verbal sounds such as audible taps cannot be detected
Solution Approach 1:
The system changes the detection parameters from speech-specific features to general acoustic features including dynamic range, periodicity, and spectral characteristics. This allows the same detection system to handle both verbal and non-verbal sounds by adjusting which parameters are analyzed and how they are interpreted.
Solution Approach 2:
The audio detection system is designed to be universal by detecting multiple types of sounds (speech, taps, claps, snaps) using a single integrated approach. The system analyzes general acoustic properties rather than being specialized for speech only, enabling it to respond to various user interactions including non-verbal sounds.
2Adaptability or versatility
If audio signals are analyzed to detect all types of sounds, then non-verbal sounds can be detected, but differentiation between user-generated sounds and ambient noise becomes difficult
Solution Approach 1:
The system applies different analysis criteria to different acoustic events. User-generated sounds like taps and claps have specific local characteristics (high dynamic range, specific periodicity patterns, particular spectral content) that distinguish them from ambient noise. The detection algorithm is tuned to recognize these localized acoustic signatures.
Solution Approach 2:
The system uses dynamic analysis of acoustic signals, examining how sound properties change over time. User-generated sounds exhibit specific dynamic patterns (sudden onset, characteristic decay, rhythmic repetition for double-taps) that differ from stationary ambient noise. The system analyzes temporal dynamics including attack, sustain, and release characteristics to differentiate sound sources.
3Measurement precision
If linear prediction coding is used to isolate audible taps, then tap detection accuracy improves, but computational complexity increases
Solution Approach 1:
The system extracts only the essential features needed for tap detection from the audio signal using linear prediction coding. Instead of analyzing the entire audio signal in detail, it extracts specific characteristics (prediction error, dynamic range, periodicity) that are most indicative of tap sounds, reducing the amount of data that needs to be processed while maintaining detection accuracy.
Solution Approach 2:
The system applies linear prediction coding selectively rather than to all audio analysis tasks. It uses this computationally intensive method specifically for detecting transient sounds like taps where high precision is needed, while potentially using simpler methods for other types of sound detection, balancing computational load with detection requirements.
Data Source
AI summary
Techniques are described for detecting and/or validating audible taps, double-taps, or other tapped audio sequences generated by a user. A predicted voice signal is subtracted from an audio signal to produce a prediction error signal. The prediction error signal is analyzed in terms of dynamics and periodicity to determine whether it contains an audio pulse. In certain cases, prediction error signals corresponding to multiple microphones are analyzed to determine position information regarding the individual taps of a double-tap. A double-tap is validated and acted upon when the individual taps originate from the same location and the taps occur within a given time period.