Continuous Audio Stream Song Detection via Segmented Acoustic Fingerprinting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio fingerprinting technologies struggle to accurately identify songs in continuous audio streams, especially in noisy environments and when multiple versions of a song exist, and they require user input or are limited by memory capacity, leading to reduced detection accuracy and inability to monitor continuous audio streams effectively.
Innovation Solution
A cloud-based system that continuously monitors Internet-based audio streams by segmenting and analyzing audio into PCM streams, generating acoustic fingerprints, and comparing them to a database to detect songs in real-time, even in noisy conditions, while managing memory usage across multiple servers to prevent disruptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single computer or server is used to monitor continuous audio streams, then song detection can be performed, but memory capacity is exceeded and temporal disruptions occur
Solution Approach 1:
The system divides the continuous audio stream into discrete audio clips of predetermined length. Each audio clip is processed independently to generate audio fingerprints, which are then matched against the database. This segmentation approach prevents memory overflow by processing manageable chunks rather than loading entire continuous streams into memory.
Solution Approach 2:
The system performs preliminary actions by pre-generating and storing audio fingerprints for all songs in the database before actual detection begins. During runtime, only fingerprint matching is performed rather than full audio analysis, significantly reducing memory requirements while maintaining detection capability.
2Ease of operation
If audio fingerprinting is performed on user-defined clips, then song identification can be achieved, but detection accuracy is less than 90% and continuous operation is impossible
Solution Approach 1:
The system enables continuous operation by automatically processing continuous audio streams without requiring user intervention to define clips. The continuous stream is segmented into overlapping or adjacent clips that are processed sequentially, allowing the system to operate indefinitely without stopping or user input while maintaining high detection accuracy through consistent fingerprint matching.
3Device complexity
If short audio segments are recorded for fingerprinting, then processing is simple, but song boundaries cannot be detected and accuracy reduces in noisy environments
Solution Approach 1:
The system dynamically adjusts the analysis by examining temporal patterns of fingerprint matches across multiple audio clips. By analyzing whether matches occur in temporally consistent sequences and using confidence scoring that considers match density and temporal coherence, the system can reliably detect song boundaries and maintain robustness against noise while keeping processing relatively simple.
Data Source
AI summary
A system and method for detecting songs in a continuous audio stream are disclosed. A detection server segments the continuous audio stream, and analyzes the audio stream to determine song candidates according to various processes disclosed herein. In one embodiment, the candidates are determined to be accurate when temporally ordered fingerprints exceed a predetermined threshold, adjacent audio stream segments are determined to have the same best song candidate, and the determined song detection has not been previously detected in the data stream within a predetermined period of time.