Sound Signal Beat Estimation Without Transition Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating time lengths of beats and measures in musical pieces from sound signals require pre-definition of transition probability models, making them complex and less efficient.
Innovation Solution
A sound signal processing method that acquires a unit time length, calculates tone feature amounts, and estimates beat and measure lengths using a first index based on the unit time length, allowing for simpler estimation without pre-defined models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If transition probability models are pre-defined for estimating beat points, tempos, and measure positions, then the estimation can be performed using existing techniques, but the system complexity increases and the process becomes less efficient
Solution Approach 1:
The patent extracts and removes the complex pre-defined transition probability models from the estimation system. Instead of using predetermined models for beat points, tempos, and measure positions, the invention directly analyzes the sound signal to extract these features, eliminating the need for complex model definitions while maintaining estimation accuracy
Solution Approach 2:
The system performs self-service by automatically determining beat points, tempos, and measure positions directly from the sound signal without relying on external pre-defined models. The estimation process serves itself by using the raw signal characteristics to derive temporal structure information, reducing system complexity
2Reliability
If pre-defined transition probability models are used for beat and measure estimation, then existing techniques can be applied, but the processing efficiency decreases
Solution Approach 1:
The patent removes the time-consuming model definition and processing steps by directly extracting beat and measure information from the sound signal. This extraction approach eliminates the intermediate modeling phase, significantly improving processing efficiency while maintaining reliable estimation results
Solution Approach 2:
The invention skips the intermediate step of using pre-defined transition probability models and directly rushes through to estimate beat points, tempos, and measure positions from the sound signal. This direct approach bypasses unnecessary processing stages, enhancing overall productivity
Data Source
AI summary
Provided is a sound signal processing method including a step of acquiring a unit time length in an input sound signal, a step of calculating a tone feature amount from the input sound signal, a step of calculating an index indicating validity of time lengths in a case where the time lengths are assumed the time lengths of one beat and one measure using the unit time length as a reference in the input sound signal with respect to the tone feature amount, and a step of estimating the time lengths of one beat and one measure using the unit time length as a reference based on the index.