On-Line Melody Extraction From Polyphonic Audio Using Harmonic Cluster Tracking

Vipul Arora, Laxmidhar Behera

    Research output: Contribution to journalArticle

    20 Citations (Scopus)

    Abstract

    Extraction of predominant melody from the musicalperformances containing various instruments is one of the mostchallenging task in the field of music information retrieval andcomputational musicology. This paper presents a novel frameworkwhich estimates predominant vocal melody in real-timeby tracking various sources with the help of harmonic clusters(combs) and then determining the predominant vocal source byusing the harmonic strength of the source. The novel on-lineharmonic comb tracking approach complies with both structuralas well as temporal constraints simultaneously. It relies uponthe strong higher harmonics for robustness against distortionof the first harmonic due to low frequency accompaniments, incontrast to the existing methods which track the pitch values. Thepredominant vocal source identification depends upon the novelidea of source dependant filtering of recognition score, whichallows the algorithm to be implemented on-line. The proposedmethod, although on-line, is shown to significantly outperformour implementation of a state-of-the-art offline method for vocalmelody extraction. Evaluations also show the reduction in octaveerror and the effectiveness of novel score filtering technique inenhancing the performance.
    LanguageEnglish
    Pages520-530
    JournalIEEE Transactions on Audio, Speech and Language Processing
    Volume21
    Issue number3
    DOIs
    Publication statusPublished - Mar 2013

    Fingerprint

    harmonics
    Information retrieval
    information retrieval
    music
    low frequencies
    evaluation
    estimates

    Cite this

    @article{124cddf715bb43cbb78328860d9d9115,
    title = "On-Line Melody Extraction From Polyphonic Audio Using Harmonic Cluster Tracking",
    abstract = "Extraction of predominant melody from the musicalperformances containing various instruments is one of the mostchallenging task in the field of music information retrieval andcomputational musicology. This paper presents a novel frameworkwhich estimates predominant vocal melody in real-timeby tracking various sources with the help of harmonic clusters(combs) and then determining the predominant vocal source byusing the harmonic strength of the source. The novel on-lineharmonic comb tracking approach complies with both structuralas well as temporal constraints simultaneously. It relies uponthe strong higher harmonics for robustness against distortionof the first harmonic due to low frequency accompaniments, incontrast to the existing methods which track the pitch values. Thepredominant vocal source identification depends upon the novelidea of source dependant filtering of recognition score, whichallows the algorithm to be implemented on-line. The proposedmethod, although on-line, is shown to significantly outperformour implementation of a state-of-the-art offline method for vocalmelody extraction. Evaluations also show the reduction in octaveerror and the effectiveness of novel score filtering technique inenhancing the performance.",
    author = "Vipul Arora and Laxmidhar Behera",
    year = "2013",
    month = "3",
    doi = "10.1109/TASL.2012.2227731",
    language = "English",
    volume = "21",
    pages = "520--530",
    journal = "IEEE Transactions on Audio, Speech and Language Processing",
    issn = "1558-7916",
    number = "3",

    }

    On-Line Melody Extraction From Polyphonic Audio Using Harmonic Cluster Tracking. / Arora, Vipul; Behera, Laxmidhar.

    In: IEEE Transactions on Audio, Speech and Language Processing, Vol. 21, No. 3, 03.2013, p. 520-530.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - On-Line Melody Extraction From Polyphonic Audio Using Harmonic Cluster Tracking

    AU - Arora, Vipul

    AU - Behera, Laxmidhar

    PY - 2013/3

    Y1 - 2013/3

    N2 - Extraction of predominant melody from the musicalperformances containing various instruments is one of the mostchallenging task in the field of music information retrieval andcomputational musicology. This paper presents a novel frameworkwhich estimates predominant vocal melody in real-timeby tracking various sources with the help of harmonic clusters(combs) and then determining the predominant vocal source byusing the harmonic strength of the source. The novel on-lineharmonic comb tracking approach complies with both structuralas well as temporal constraints simultaneously. It relies uponthe strong higher harmonics for robustness against distortionof the first harmonic due to low frequency accompaniments, incontrast to the existing methods which track the pitch values. Thepredominant vocal source identification depends upon the novelidea of source dependant filtering of recognition score, whichallows the algorithm to be implemented on-line. The proposedmethod, although on-line, is shown to significantly outperformour implementation of a state-of-the-art offline method for vocalmelody extraction. Evaluations also show the reduction in octaveerror and the effectiveness of novel score filtering technique inenhancing the performance.

    AB - Extraction of predominant melody from the musicalperformances containing various instruments is one of the mostchallenging task in the field of music information retrieval andcomputational musicology. This paper presents a novel frameworkwhich estimates predominant vocal melody in real-timeby tracking various sources with the help of harmonic clusters(combs) and then determining the predominant vocal source byusing the harmonic strength of the source. The novel on-lineharmonic comb tracking approach complies with both structuralas well as temporal constraints simultaneously. It relies uponthe strong higher harmonics for robustness against distortionof the first harmonic due to low frequency accompaniments, incontrast to the existing methods which track the pitch values. Thepredominant vocal source identification depends upon the novelidea of source dependant filtering of recognition score, whichallows the algorithm to be implemented on-line. The proposedmethod, although on-line, is shown to significantly outperformour implementation of a state-of-the-art offline method for vocalmelody extraction. Evaluations also show the reduction in octaveerror and the effectiveness of novel score filtering technique inenhancing the performance.

    U2 - 10.1109/TASL.2012.2227731

    DO - 10.1109/TASL.2012.2227731

    M3 - Article

    VL - 21

    SP - 520

    EP - 530

    JO - IEEE Transactions on Audio, Speech and Language Processing

    T2 - IEEE Transactions on Audio, Speech and Language Processing

    JF - IEEE Transactions on Audio, Speech and Language Processing

    SN - 1558-7916

    IS - 3

    ER -