Filters








32 Hits in 4.7 sec

A case study of deep enculturation and sensorimotor synchronization to real music

Olof Misgeld, Torbjörn L Gulz, Jūra Miniotaitė, Andre Holzapfel
2021 Zenodo  
The identified differences between groups are related to the metrical structures inherent to the two musical styles, such as non-isochronicity of the beat, and differences between the groups document the  ...  Instead of choosing two apparently remote cultural groups, we selected two groups of musicians that share cultural backgrounds, but that differ regarding the music style they specialize in.  ...  Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 Figure 2 . 2 Figure 2.  ... 
doi:10.5281/zenodo.5624537 fatcat:z36f7aa3bvfrrjjsguugluni3q

On the Integration of Language Models into Sequence to Sequence Architectures for Handwritten Music Recognition

Pau Torras, Arnau Baró, Lei Kang, Alicia Fornés
2021 Zenodo  
Despite the latest advances in Deep Learning, the recognition of handwritten music scores is still a challenging endeavour.  ...  Indeed, the ambiguous nature of handwriting, the non-standard musical notation employed by composers of the time and the decaying state of old paper make these scores remarkably difficult to read, sometimes  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.  ... 
doi:10.5281/zenodo.5624451 fatcat:hr7x55idwfaphnj5dpcllmlbui

ADTOF: A large dataset of non-synthetic music for automatic drum transcription

Mickael Zehren, Marco Alunno, Paolo Bientinesi
2021 Zenodo  
When used to train a popular DTM model, the dataset yields a performance that matches that of the state-of-the-art for DTM, thus demonstrating the quality of the annotations.  ...  This dataset contains real-world music, is manually annotated, and is about two orders of magnitude larger than any other non-synthetic dataset, making it a prime candidate for training purposes.  ...  Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 Table 1 . 1 List of datasets for DTM.  ... 
doi:10.5281/zenodo.5624527 fatcat:zpjqkio4wbb2xc7ovdk4lhbg4a

Cross-cultural Mood Perception in Pop Songs and its Alignment with Mood Detection Algorithms

Harin Lee, Frank Höger, Marc Schönwiesner, Minsu Park, Nori Jacoby
2021 Zenodo  
Analyzing 166 participants' responses from Brazil, South Korea, and the US, we examined the similarity between the ratings of nine categories of perceived moods in music and estimated their alignment with  ...  Do people from different cultural backgrounds perceive the mood in music the same way?  ...  ISMIR Conference, Online, November 7-12, 2021  ... 
doi:10.5281/zenodo.5625680 fatcat:emirislqmnanhddh6ki6fj5ulm

Neural Waveshaping Synthesis

Ben Hayes, Charalampos Saitis, Gyorgy Fazekas
2021 Zenodo  
We paired the NEWT with a differentiable noise synthesiser and reverb and found it capable of generating realistic musical instrument performances with only 260k total model parameters, conditioned on  ...  ) for efficient CPU inference.  ...  ACKNOWLEDGEMENTS We would like to thank the anonymous reviewers at ISMIR for their thoughtful comments.  ... 
doi:10.5281/zenodo.5624613 fatcat:y5ux347cwjhv5oi3mttbyzgsd4

SpecTNT: a Time-Frequency Transformer for Music Audio

Wei-Tsung Lu, Ju-Chiang Wang, Minz Won, Keunwoo Choi, Xuchen Song
2021 Zenodo  
Then, a temporal Transformer processes the TEs to exchange information across the time axis. By stacking the SpecTNT blocks, we build the SpecTNT model to learn the representation for music signals.  ...  In experiments, SpecTNT demonstrates state-of-the-art performance in music tagging and vocal melody extraction, and shows competitive performance for chord recognition.  ...  Out of the whole Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 track, we use 400 frames as an input.  ... 
doi:10.5281/zenodo.5624503 fatcat:ru5bho3jwvefvize5pxo7gcyni

Audiovisual Singing Voice Separation

Bochen Li, Yuxuan Wang, Zhiyao Duan
2021 Transactions of the International Society for Music Information Retrieval  
Query by video: Crossmodal Late-Breaking Demo, International Society for Music music retrieval. In Proceedings of the International Information Retrieval Conference (ISMIR).  ...  are subscribers from the International of visual information is only processed by the Conv3D Society for Music Information Retrieval (ISMIR) structure.  ... 
doi:10.5334/tismir.108 fatcat:5k2cd26tufbi7mv7lx6kvlomee

A Differentiable Cost Measure for Intonation Processing in Polyphonic Music

Simon J Schwär, Sebastian Rosenzweig, Meinard Müller
2021 Zenodo  
Intonation is the process of choosing an appropriate pitch for a given note in a musical performance.  ...  In an experiment, we demonstrate the potential of our approach for the task of intonation adaptation of amateur choral music using recordings from a publicly available multitrack dataset.  ...  The International Audio Laboratories Erlangen are a joint institution of the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) and Fraunhofer Institute for Integrated Circuits IIS.  ... 
doi:10.5281/zenodo.5624601 fatcat:ggmjxt437jg4rct64h2dpcvw6a

MuseBERT: Pre-training Music Representation for Music Understanding and Controllable Generation

Ziyu Wang, Gus Xia
2021 Zenodo  
Experiment shows that the pre-trained model outperforms the baselines in terms of reconstruction likelihood and generation quality.  ...  , 2) the pre-trained MuseBERT is not merely a language model, but also a controllable music generator, and 3) MuseBERT gives birth to various downstream music generation and analysis tasks with practical  ...  Proceedings of the 22nd ISMIR Conference, Online, Musicalion: https://www.musicalion.com. Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021  ... 
doi:10.5281/zenodo.5624387 fatcat:uum2nyfrkreuxi5t356tqknscq

Semi-supervised Music Tagging Transformer

Minz Won, Keunwoo Choi, Xavier Serra
2021 Zenodo  
Through a careful model assessment, we first show that the proposed architecture outperforms the previous state-of-the-art music tagging models that are based on convolutional neural networks under a supervised  ...  To our best knowledge, this is the first attempt to utilize the entire audio of the million song dataset.  ...  Our back end Transformer architecture is nearly iden-Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 tical to the previous works [20, 21] except for the number of parameters and  ... 
doi:10.5281/zenodo.5624405 fatcat:jxi3c6edbrelplsmbayqjo6e24

MINGUS: Melodic Improvisation Neural Generator Using Seq2Seq

Vincenzo Madaghiele, Pasquale Lisena, Raphael Troncy
2021 Zenodo  
The obtained results are comparable with the state of the art of music generation with neural models, with particularly good performances on jazz music.  ...  MINGUS relies on two dedicated embedding models (respectively for pitch and duration) and exploits in prediction features such as chords (current and following), bass line, position inside the measure.  ...  Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 Model Architecture MINGUS is structured as two parallel transformer models with the same structure, respectively predicting pitch  ... 
doi:10.5281/zenodo.5625684 fatcat:qhefigcrqzgxvoepnyag4vpxpy

Pitch-Informed Instrument Assignment using a Deep Convolutional Network with Multiple Kernel Shapes

Carlos Lordelo, Emmanouil Benetos, Simon Dixon, Sven Ahlbäck
2021 Zenodo  
We also include ablation studies investigating the effects of the use of multiple kernel shapes and comparing different input representations for the audio and the note-related information.  ...  Experiments on the MusicNet dataset using 7 instrument classes show that our approach is able to achieve an average F-score of 0.904 when the original multi-pitch annotations are used as the pitch information  ...  Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 Figure 4 . 4 Figure 4. Internal structure of a multi-branch conv. stage.  ... 
doi:10.5281/zenodo.5625681 fatcat:ia4vngvnqrfhfjg6tr4mhkyxna

Leveraging Hierarchical Structures for Few-Shot Musical Instrument Recognition

Hugo F Flores Garcia, Aldo Aguilar, Ethan Manilow, Bryan Pardo
2021 Zenodo  
We apply a hierarchical loss function to the training of prototypical networks, combined with a method to aggregate prototypes hierarchically, mirroring the structure of a predefined musical instrument  ...  Deep learning work on musical instrument recognition has generally focused on instrument classes for which we have abundant data.  ...  Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021 https://github.com/hugofloresgarcia/music-trees Proceedings of the 22nd ISMIR Conference, Online, November 7-12, 2021  ... 
doi:10.5281/zenodo.5624615 fatcat:kwpblujcsffbbckacgrkurqzp4

Cosine Contours: a Multipurpose Representation for Melodies

Bas Cornelissen, Willem Zuidema, John Ashley Burgoyne
2021 Zenodo  
independent of the specifics of the data sets for which it is used.  ...  The motivation for this approach is twofold: (1) it approximates a maximally informative contour representation (capturing most of the variation in as few dimensions as possible), but (2) it is nevertheless  ...  ACKNOWLEDGEMENTS We would like to thank Henkjan Honing and Marianne de Heer Kloots for their feedback on the manuscript.  ... 
doi:10.5281/zenodo.5624530 fatcat:4nefzhjtn5bqdoolu4nu5hjs5u

Toward an Understanding of Lyrics-viewing Behavior While Listening to Music on a Smartphone

Kosetsu Tsukuda, Masahiro Hamasaki, Masataka Goto
2021 Zenodo  
Better understanding of lyrics viewing behavior would be beneficial for both researchers and music streaming platforms to improve their lyrics-based systems.  ...  To answer "how," we analyze over 23 million lyrics request logs sent from the smartphone application of a music streaming service.  ...  The authors would like to extend their appreciation to SyncPower Corporation for providing the lyrics request logs.  ... 
doi:10.5281/zenodo.5624632 fatcat:2w5iwpamxbccfa4rufzv5rarly
« Previous Showing results 1 — 15 out of 32 results