Music Information Retrieval Using Hit Count Vector
نویسنده
چکیده
Roughly speaking, music can be easily and cheaply accessed via Internet, but at the same time, finding certain music is still a difficult task. In this paper, a music retrieval system based on frequency descriptors is presented. The work flow of proposed system passes through two main phases: (i) the enrollment phase and (ii) the retrieval phase. In both phases, two models are presented (i.e., preprocessing and feature extraction module). Transformation is an important stage in audio processing application; it is applied to convert the audio signal from time domain to frequency domain. Two types of frequency transformation have been used in this work: (i) Fourier and (ii) discrete cosine transform (DCT). The proposed method implies the application of several processing stages on the input melody files; these stages are considered as parts of the preparation operation. The task of music features extraction for retrieval purpose is a challenging problem. An optimization algorithm is developed to determine the proper threshold value that can filter-in a set of the most energetic and frequent frequencies existing in the spectra, such that their number should be sufficient. This set of frequencies is used to establish the discriminating features, called hit count array.In the enrollment phase, the extracted hit count array is stored in the system database as a template to be used in the retrieval process phase. Euclidean and City Block similarity metrics have been used to make a decision in the retrieval stage. For performance evaluate, several performance measures (like, sensitivity, specificity, recall, precision, accuracy) have been used. The system was tested using a dataset consists of 50 audio files; each has the specifications (11025 sampling rate, 8-bit resolution and mono channel). The length of taken Melodies is about one minute; they saved in a database. Each melody sample is partitioned into 30 frames; each frame is about two seconds. The achieved retrieval results indicated high recognition rate (99.45%) when using the input query length equal to 35 seconds.
منابع مشابه
Prototyping a Vibrato-Aware Query-By-Humming (QBH) Music Information Retrieval System for Mobile Communication Devices: Case of Chromatic Harmonica
Background and Aim: The current research aims at prototyping query-by-humming music information retrieval systems for smart phones. Methods: This multi-method research follows simulation technique from mixed models of the operations research methodology, and the documentary research method, simultaneously. Two chromatic harmonica albums comprised the research population. To achieve the purpose ...
متن کاملPhoneme Recognition in Popular Music
Automatic lyrics synchronization for karaoke applications is a major challenge in the field of music information retrieval. An important pre-requisite in order to precisely synchronize the music and corresponding text is the detection of single phonemes in the vocal part of polyphonic music. This paper describes a system, which detects the phonemes based on a state-of-the-art audio information ...
متن کاملشناسایی خودکار سبک موسیقی
Nowadays, automatic analysis of music signals has gained a considerable importance due to the growing amount of music data found on the Web. Music genre classification is one of the interesting research areas in music information retrieval systems. In this paper several techniques were implemented and evaluated for music genre classification including feature extraction, feature selection and m...
متن کاملEffectiveness of Note Duration Information for Music Retrieval
Content-based music information retrieval uses features extracted from music to answer queries. For melodic queries, the two main features are the pitch and duration of notes. The note pitch feature has been well researched whereas duration has not been fully explored. In this paper, we discuss how the note duration feature can be used to alter music retrieval effectiveness. Notes are represent...
متن کاملEmotion Based Information Retrieval System
Music emotion plays an important role in music retrieval, mood detection and other music-related applications. Many issues for music emotion recognition have been addressed by different disciplines such as physiology, psychology, cognitive science and musicology. We present a support vector regression (SVR) based Music Information Retrieval System (Emotion based). We have chosen the “Raga” para...
متن کامل