Machine Learning Sound to Vibration Waveform Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating vibration waveform data for vibration devices are time-consuming and labor-intensive, with a lack of effective automated solutions that replicate manually created waveforms.
Innovation Solution
An information processing apparatus and method that uses machine learning to convert sound waveform data into vibration waveform data, employing sound data acquisition, teacher vibration data acquisition, and machine learning to generate learned model data, which analyzes frequency spectra for input features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If vibration waveform data is manually created by a person, then the quality and accuracy of the vibration data is improved, but the time and effort required increases significantly
Solution Approach 1:
The patent uses machine learning to automatically copy and transform sound waveform data into vibration waveform data, replicating the quality of manually created vibration data without the time investment. The learned model data captures the transformation patterns from sound to vibration waveforms, enabling automatic generation of high-quality vibration data that mimics manual creation results.
Solution Approach 2:
The patent replaces the manual mechanical process of creating vibration waveform data with an automated machine learning system. The machine learning section automatically performs the transformation from sound to vibration waveforms using learned model data, substituting human manual work with an automated computational process that maintains quality while reducing time and effort.
2Productivity
If automated generation of vibration waveform data from sound waveform data is implemented, then the time and effort required is reduced, but the similarity to manually created vibration data is insufficient
Solution Approach 1:
The patent performs preliminary action by pre-training a machine learning model using teacher vibration data that was manually created. The machine learning section learns the transformation patterns from sound to vibration waveforms by studying examples of manually created vibration data, thereby capturing the nuances and quality characteristics of manual creation in advance. This preliminary learning enables the automated system to generate vibration data that closely resembles manually created data.
Solution Approach 2:
The patent uses parameter changes by analyzing frequency spectrum values of sound data as input feature amounts and transforming them into corresponding vibration waveform parameters. The machine learning model learns the relationship between sound frequency characteristics and vibration waveform characteristics, automatically adjusting parameters to generate vibration data that matches the quality and characteristics of manually created vibration data.
3Measurement precision
If frequency spectrum analysis is used as input feature amount for machine learning, then the conversion accuracy from sound to vibration waveform is improved, but the processing complexity increases
Solution Approach 1:
The patent extracts the essential frequency spectrum values from sound data as input feature amounts for the machine learning model. By taking out and focusing on the frequency spectrum characteristics, the system captures the most relevant information for converting sound to vibration waveforms, improving conversion accuracy while avoiding unnecessary processing of unrelated sound data characteristics.
Data Source
AI summary
An information processing apparatus is configured to acquire sound data, acquire, as teacher vibration data, vibration data that is created on the basis of the sound data and that is used to cause a vibration device to vibrate, and execute machine learning by using the sound data and the teacher vibration data, to generate learned model data that is used to convert an input sound waveform into an output vibration waveform. The information processing apparatus executes the machine learning by using a value obtained by analysis of a frequency spectrum of the sound data as an input feature amount.