A Robust Audio Fingerprinting Using a New Hashing Method

Abstract

To enhance the tracking performance of illegal audio copies, we introduce a robust audio fingerprinting method against various attacks in this paper. Most audio fingerprints consist of the information in the frequency band of audio. These fingerprinting methods may lose the uniqueness of the audio fingerprint by irregular movement such as an attack with pitch value changes. The proposed fingerprint method in a fundamental frequency band makes up for the weakness of existing methods generated from frequency domain. Using the geometrical property of the proposed method, a new hashing method is employed in the similarity calculation process to compare the audio contents. In order to prove the validity of proposed algorithm, we experiment for six environments such as tempo, pitch, speed modification, noise addition, low pass filter and high pass filter. The proposed method shows the highest level of performance in most experimental environments. Especially, with respect to the tempo, pitch, and speed manipulation experiments, the proposed method archives the precision rate of the range from 95 to 100% according to the degree of manipulation, which compares favorably with the precision rate obtained by traditional approaches, and yields a precision rate between 85 and 100% in noise addition and filtering experiments.

Anime Audio Retrieval Based on Audio Separation and Feature Recognition

“…Tey conducted experiments in six diferent environments, including rhythm, pitch, speed changes, and noise addition, and employed a novel hashing method for audio content comparison in the similarity calculation process, leading to a signifcant improvement in accuracy. With the development of deep learning, there have also been deep learning-based audio fngerprinting methods [22,23].…”

Accuracy comparisons of fingerprint based song recognition approaches using very high granularity

“…Moreover, memory requirements to store the data structures used to recognize the songs are very large. In [ 25 ], the fundamental frequency components extracted from the audio were matched with the frame-fundamental frequency domain and used to compose what the authors call fundamental frequency map (FFMAP). Authors employed also a new hashing method named spatial adaptive hashing (SAH) in the similarity calculation process, to compare the audio contents.…”