Background Music Removal Based on Cepstrum Transformation for Popular Singer Identification

Tsai, Wei-Ho; Lin, Hao-Ping

doi:10.1109/tasl.2010.2087752

Cited by 22 publications

(15 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The superimposition was done carefully, such that the resulting accompanied singing voices are in rhythm. The definition of SAR refers to [11]. Those superimposed mixtures serve as testing data set and partly as training data set for SAR-specific MLP.…”

Section: Methodsmentioning

confidence: 99%

“…Tsai et.al [11] proposed a background accompaniment re moval approach for singer identification exploiting the un derling relationships between a solo singing voice and the accompanied sound in a cepstrum. They obtained the cep strums of the singing voices by transforming the cepstrums of accompanied singing.…”

Section: Comparison With a Related Systemmentioning

confidence: 99%

“…Some researches were dedicated to reduce or remove the negative influences caused by instrumental accompaniment in modeling the characteris tics of a singing voice [3]. Tsai et al [11] characterized a singing voice by "transforming " the cepstrum of an accom panied singing voice into a solo singing voice cepstrum. The transformed cepstral coefficients are regarded as the features which capture the singer's characteristics.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Automatic singer identification using missing feature methods

Liu

2013

2013 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

This study for singer identification of mono popular music is in two stages. In the first stage, computational auditory scene analysis (CASA) is exploited to segregate singing voice units. For each frame, the estimated binary T-F mask indi cates the time-frequency (T-F) units dominated by singing voice which are considered reliable, and other units are un reliable or missing. Thus the spectrum is incomplete. In the second stage, two missing feature methods, reconstruc tion and marginalization are used to identify the singer by dealing with the incomplete spectrum data. In the reconstruc tion module, the complete spectrum is first reconstructed and then converted to obtain the Gammatone frequency cepstral coefficients (GFCCs), which are further used to identify the singer. In the marginalization module, the probabilities of the singer's voice are computed on the basis of only the re liable components. We find that the reconstruction module outperforms the marginalization module, while both modules have significantly good performances, especially at signal-to accompaniment ratios (SARs) of 0 dB and -3 dB, in contrast to other system. Index Terms-Singer identification, missing feature, computational auditory scene analysis (CASA).

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Comparison With a Related Systemmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic singer identification using missing feature methods

Liu

2013

2013 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

show abstract

“…EM algorithm is high for two major reasons as similar to other kernel based methods, it has to calculate kernel function for each samplepair over training set and in order to obtain the largest eigen value [9]. There are two main reasons for using the SVM in audio classification.…”

Section: B) Gaussian Mixture Modelmentioning

confidence: 99%

Survey on

2016

IJSR

View full text Add to dashboard Cite

Singing voice separation from music is a kind of speech separation and it is a big challenge in many applications.Humanauditory system has a good ability of effectively focusing on sound in the surrounding. Most audio signals arefrom the mixing of several sound sources. Separation of singing voice from music has wide range of application suchas lyrics recognition, alignment, singer identification, and music information retrieval. Music accompaniment that isoften non-stationary & harmonic. Basically, audio signal is time frequency segments of singing voice. An audio signalclassification system should be able to categorize different audio format like speech, background noise, and musicalgenres, singer identification, karaoke etc. In this paper, discuss about different separation technique and classifier which are used for singing voice separation from music.

show abstract

“…Indeed, the vocal timbre and singing style can influence people's decision on which songs to listen to. In fact, several music information retrieval (MIR) systems based on vocal timbre similarity have been proposed [1][2][3][4][5]. When people listen to singing voices, they can feel that different vocal timbres and singing styles share some factors that characterize their timbres and styles.…”

Section: Introductionmentioning

confidence: 99%

Vocal timbre analysis using latent Dirichlet allocation and cross-gender vocal timbre similarity

Nakano

Yoshii

Goto

2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

This paper presents a vocal timbre analysis method based on topic modeling using latent Dirichlet allocation (LDA). Although many works have focused on analyzing characteristics of singing voices, none have dealt with "latent" characteristics (topics) of vocal timbre, which are shared by multiple singing voices. In the work described in this paper, we first automatically extracted vocal timbre features from polyphonic musical audio signals including vocal sounds. The extracted features were used as observed data, and mixing weights of multiple topics were estimated by LDA. Finally, the semantics of each topic were visualized by using a word-cloud-based approach. Experimental results for a singer identification task using 36 songs sung by 12 singers showed that our method achieved a mean reciprocal rank of 0.86. We also proposed a method for estimating cross-gender vocal timbre similarity by generating pitch-shifted (frequency-warped) signals of every singing voice. Experimental results for a cross-gender singer retrieval task showed that our method discovered interesting similar pitch-shifted singers.

show abstract

Background Music Removal Based on Cepstrum Transformation for Popular Singer Identification

Cited by 22 publications

References 23 publications

Automatic singer identification using missing feature methods

Automatic singer identification using missing feature methods

Survey on

Vocal timbre analysis using latent Dirichlet allocation and cross-gender vocal timbre similarity

Contact Info

Product

Resources

About