Audio Melody Detection via Beat Segmentation and Pitch Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for recognizing melodies in audio signals, especially in cases of poor singing or humming without explicit lyrics, face challenges in accuracy due to pitch inaccuracies and lack of applicable database matching.
Innovation Solution
A method that divides audio signals into segments based on beats, detects pitch frequencies, estimates pitch values, determines pitch names, and calculates musical scales to accurately identify melodies, even in cases of off-key singing, using techniques like Short-Time Fourier Transform and toning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional voice recognition and lyric matching methods are used to recognize melodies, then the system can identify songs with clear lyrics, but it fails to accurately recognize melodies when users sing with pitch inaccuracies or hum without explicit lyrics
Solution Approach 1:
The patent transforms the melody recognition problem from lyric-based matching to pitch-frequency-based detection. By changing the detection parameter from textual lyrics to acoustic pitch values, the system can handle off-key singing and humming without requiring exact lyric matches or perfect pitch accuracy.
Solution Approach 2:
The patent replaces the conventional lyric-matching mechanism with a pitch detection mechanism using Short-Time Fourier Transform. This substitution allows the system to directly analyze audio frequency characteristics rather than relying on text recognition and database matching, enabling accurate melody detection even when lyrics are unclear or pitch is inaccurate.
2Measurement precision
If the system divides audio signals into segments and performs detailed pitch frequency detection on each frame, then melody detection accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The patent divides the audio signal into multiple frames and further segments each frame into audio sub-segments for pitch detection. This segmentation allows the system to process complex audio data in manageable portions, improving pitch detection accuracy while controlling computational complexity through localized analysis of smaller signal portions.
Solution Approach 2:
The patent performs pitch frequency detection on selected frames and audio sub-segments rather than processing every single frame in detail. By applying partial action (processing only critical segments) or excessive action (processing more segments than strictly necessary for redundancy), the system achieves high accuracy while managing computational load through selective detailed analysis.
Data Source
AI summary
A method for detecting a melody of an audio signal, including: dividing the audio signal into a plurality of audio segments based on a beat, detecting a pitch frequency of each frame of audio sub-signal in each of the audio segments, and estimating a pitch value of each of the audio segments based on the pitch frequency; determining a pitch name corresponding to each of the audio segments based on a frequency range of the pitch value; acquiring a musical scale of the audio signal by estimating a tonality of the audio signal based on the pitch name of each of the audio segments; and determining a melody of the audio signal based on a frequency interval of the pitch value of each of the audio segments in the musical scale.