2,061 Hits in 6.9 sec

Novelty Detection Based On Spectral Similarity Of Songs

Arthur Flexer, Elias Pampalk, Gerhard Widmer
2005 Zenodo  
ACKNOWLEDGEMENTS Parts of the MA Toolbox [13] and the Netlab Toolbox 2 have been used for this work. This research was supported by the EU project FP6-507142 SIMAC 3 .  ...  The Austrian Research Institute for Artificial Intelligence is supported by the Austrian Federal Ministry of Education, Science and Culture and the Austrian Federal Ministry for Transport, Innovation and  ...  We will present two methods for novelty detection based on spectral similarity of songs and evaluate them within a genre classification context (see e.g. [2] ).  ... 
doi:10.5281/zenodo.1416503 fatcat:4lo5gkmquvffnnyjzzj565ooqy

Musical audio semantic segmentation exploiting analysis of prominent spectral energy peaks and multi-feature refinement

P. Romano, G. Prandi, A. Sarti, S. Tubaro
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
Experimental results on a set of 58 songs show that our algorithm is able to attain good semantic segmentation just after the first step, with a precision of 64% and a recall of 96%.  ...  In this step the minimum average recall value of 92% is obtained.  ...  To perform segmentation in the first step, we present a new segmentation algorithm based on spectral energy texture change detection.  ... 
doi:10.1109/icassp.2009.4959996 dblp:conf/icassp/RomanoPST09 fatcat:r6darqra3vbshmn25u6l3zyogu

A Causal Rhythm Grouping [chapter]

Kristoffer Jensen
2005 Lecture Notes in Computer Science  
The method is based on a hierarchical model; first a features is measured from the audio, then a measure of rhythm is calculated from the feature (the rhythmogram), the diagonal of a self-similarity matrix  ...  is calculated from the rhythmogram, and finally the segment boundaries are found on a smoothed novelty measure, calculated from the diagonal of the self-similarity matrix.  ...  Later Goto and Muraoka [12] developed a system to perform beat tracking independent of drum sounds, based on detection of chord changes.  ... 
doi:10.1007/978-3-540-31807-1_6 fatcat:sxjaia6wjvc7llweghsb6yq5ce

Segmentation And Timbre Similarity In Electronic Dance Music

Bruno Rocha, Niels Bogaards, Aline Honingh
2013 Proceedings of the SMC Conferences  
This research has been funded by the Center for Digital Humanities ( and the Netherlands foundation of Scientific Research (NWO), grant no. 639.021.126.  ...  Song: "Insomnia (Monster Mix)" by Faithless. Figure 2 . 2 Novelty detection. (a) Similarity matrix with kernel size of approximately 30 seconds, (b) Novelty curve. Song: "& Down" by Boys Noize.  ...  As the novelty detection is based on textural changes and the timbres of articulative sounds are frequently quite distinct from the neighbours, novelty peaks are detected when these sounds occur.  ... 
doi:10.5281/zenodo.850376 fatcat:xdnly5usejgqfaluhymxj3boyq

Exploiting global features for tempo octave correction

Hendrik Schreiber, Meinard Muller
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
While using a standard pulse detection technique, for octave error correction, we exploit a simple relationship between a single global feature, average spectral novelty, and listener perception of musical  ...  Nevertheless, it outperforms most existing tempo estimation methods and is on par with the best-performing ones.  ...  From these features, the mean spectral novelty (SNM) turned out to be the most successful one.  ... 
doi:10.1109/icassp.2014.6853674 dblp:conf/icassp/SchreiberM14 fatcat:3j3gckg6u5gs3khheo5q2wlt7q

An Automatic Segmentation Method of Popular Music Based on SVM and Self-similarity [chapter]

Feng Li, You You, Yuqin Lu, YuQing Pan
2015 Lecture Notes in Computer Science  
Segmentation boundary detection is one of the key technologies in lots of conventional algorithms.  ...  Finally, self-similarity detecting algorithm is introduced to refine our segmentation results in the vicinity of potential points.  ...  In comparison with the method based on self-similarity deployed to detect the music segmentation boundaries without pre-process procedure, our method use the self-similarity detect algorithm only on the  ... 
doi:10.1007/978-3-319-15554-8_2 fatcat:iozbhh3obfczldobc3uoimtzrm

An Integrated Approach To Music Boundary Detection

Min-Yian Su, Yi-Hsuan Yang, Yu-Ching Lin, Homer H. Chen
2009 Zenodo  
The authors would like to thank the anonymous reviewers for valuable comments that greatly improved the quality of this paper.  ...  ACKNOWLEDGEMENT This work was supported by the National Science Council of Taiwan under the contract number NSC 97-2221-E-002-111-MY3.  ...  Unsupervised Estimation As for the unsupervised part, we construct three similarity matrices based on the kinds of low-level features and detect the peaks of the mean of the novelty scores from these matrices  ... 
doi:10.5281/zenodo.1417584 fatcat:bhnnlvwabfc3pforl3djhenhza

Media segmentation using self-similarity decomposition

Jonathan T. Foote, Matthew L. Cooper, Minerva M. Yeung, Rainer W. Lienhart, Chung-Sheng Li
2003 Storage and Retrieval for Media Databases 2003  
The summaries can be customized for various applications based on the structure of the original music.  ...  In the first step, spectral data is used to construct a similarity matrix calculated from inter-frame spectral similarity.  ...  The right panel shows the novelty score computed from the similarity matrix in Figure 1 . Large peaks are detected in the resulting time-indexed correlation and labelled as segment boundaries.  ... 
doi:10.1117/12.476302 dblp:conf/spieSR/FooteC03 fatcat:krvum7ootnhkdphodr4ttuctze

A Simple Fusion Method Of State And Sequence Segmentation For Music Structure Discovery

Florian Kaiser, Geoffroy Peeters
2013 Zenodo  
CASE STUDY In this section we illustrate on a real signal (the song "One Vision" by Queen) our method for segment detection based on late-fusion of repetition and novely-based segmentation.  ...  Three novelty curves are computed for all three kernels on a SSM computed on timbre-related features (MFCCs, Spectral Centroid, Spread, Skewness and Spectral Flatness).  ... 
doi:10.5281/zenodo.1416046 fatcat:p6rdhe6qxnh43mb6ajah4swdgy

AutoMashUpper: Automatic Creation of Multi-Song Music Mashups

Matthew E. P. Davies, Philippe Hamel, Kazuyoshi Yoshii, Masataka Goto
2014 IEEE/ACM Transactions on Audio Speech and Language Processing  
The principal novelty in our approach centres on the determination of how elements of songs can be made fit together using key transposition and tempo modification, rather than based on their unaltered  ...  We define mashability in terms of harmonic and rhythmic similarity and a measure of spectral balance.  ...  Then, to extract phrase section boundaries, we follow the self-similarity and novelty function based approach of Foote [25] for structural segmentation.  ... 
doi:10.1109/taslp.2014.2347135 fatcat:33o4gebnp5banmwdkyhlhh2o7m

Evolutionarily conserved coding properties favour the neuronal representation of heterospecific signals of a sympatric katydid species

Konstantinos Kostarakos, Heiner Römer
2018 Journal of Comparative Physiology A. Sensory, neural, and behavioral physiology  
Some interneurons of the trilling species respond exclusively to the heterospecific chirp due to selective, low-frequency tuning and "novelty detection".  ...  Instead we suggest that chirpers evolved an additional, 2-kHz component in their song and exploited pre-existing neuronal properties for detecting their song under masking noise.  ...  The experiments complied with the 'Principles of animal care' given in publication no. 86 23, revised in 1985, from the National Institutes of Health and with the current laws in Austria.  ... 
doi:10.1007/s00359-018-1282-0 pmid:30225517 pmcid:PMC6182671 fatcat:pkc2p7mfmbcwhjel3nbfgpcmry

Towards Automatic Music Structural Analysis Identifying Characteristic Within-Song Excerpts in Popular Music

Bee Suan Ong, Xavier Serra
2005 Zenodo  
A database of 54 audio files (The Beatles' song) is used for the evaluation of the proposed approach on a mainstream popular music collection.  ...  In this research work, we focus our investigation on two areas that are part of audiobased music structural analysis.  ...  Given that novelty detection is based on the correlation process, the width of the kernel affects the resolution of the detection outcome.  ... 
doi:10.5281/zenodo.3739314 fatcat:ptqosukiyvaxvdxr7sncrbtuhq

Mind the beat: detecting audio onsets from EEG recordings of music listening [article]

Ashvala Vinay, Alexander Lerch, Grace Leslie
2021 arXiv   pre-print
We compare our RNN network to both the standard spectral-flux based novelty function and the FCN.  ...  We generate a sequence of onset labels for the songs in our dataset and trained neural networks (a fully connected network (FCN) and a recurrent neural network (RNN)) to parse one second windows of input  ...  RESULTS AND DISCUSSION Simply applying traditional music oriented methods to EEG data did not perform well on this task -peak-picking a spectral flux-based novelty function yielded an average Fmeasure  ... 
arXiv:2102.06393v1 fatcat:rtpbvno4q5euveqqekpgmg7x64

To catch a chorus, verse, intro, or anything else: Analyzing a song with structural functions [article]

Ju-Chiang Wang and Yun-Ning Hung and Jordan B. L. Smith
2022 arXiv   pre-print
We also propose to use a spectral-temporal Transformer-based model, called SpecTNT, which can be trained with an additional connectionist temporal localization (CTL) loss.  ...  chorus-detection and boundary-detection methods at detecting choruses and boundaries, respectively.  ...  detecting repeating sequences or relative novelty changes.  ... 
arXiv:2205.14700v1 fatcat:zm53xuzcjnd6vbs7cqlhdshkxy

"The way it Sounds": timbre models for analysis and retrieval of music signals

J.-J. Aucouturier, F. Pachet, M. Sandler
2005 IEEE transactions on multimedia  
While there has been a large quantity of research done to model the timbre of individual instruments, little work has been done to analyze "real world" timbre mixtures such as the ones found in popular  ...  Notably, we describe their applications for music similarity, segmentation and pattern induction. and developed the vision that metadata can greatly enhance the musical experience in all its dimensions  ...  Novelty-based algorithms have the disadvantage of not providing any "understanding" of the segmentation: They detect boundaries, but do not compare and label the resulting segments.  ... 
doi:10.1109/tmm.2005.858380 fatcat:ltg6rarwtreuxmiihogi2jtxky
« Previous Showing results 1 — 15 out of 2,061 results