3 Hits in 2.0 sec

Beat Tracking With A Cepstroid Invariant Neural Network

Anders Elowsson
2016 Zenodo  
"Beat Tracking with a Cepstroid Invariant Neural Network", 17th International Society for Music Information Retrieval Conference, 2016.  ...  SUMMARY & CONCLUSIONS We have presented a novel beat tracking and tempo estimation system that uses a cepstroid invariant neural network.  ... 
doi:10.5281/zenodo.1416054 fatcat:2wvpkk3evnf5jdumbtb5upzwga

Tempo-Invariant Processing of Rhythm with Convolutional Neural Networks [article]

Anders Elowsson
2018 arXiv   pre-print
This paper describes how a log-frequency representation of rhythm-related activations instead can promote tempo invariance when processed with convolutional neural networks.  ...  Several recent systems for tempo estimation, beat tracking, and downbeat tracking have therefore sought to process rhythm in a tempo-invariant way, often by sampling input vectors according to a precomputed  ...  Time-invariant neural networks can of course also be implemented as feedforward neural networks, shifted and applied to each frame (Elowsson, 2016 ).  ... 
arXiv:1804.08167v2 fatcat:22zbyym6aveoxf77uczarsdnu4

Convolutional Neural Networks For Real-Time Beat Tracking: A Dancing Robot Application

Aggelos Gkiokas, Vassilios Katsouros
2017 Zenodo  
In [17] , under the assumption that humans perceive the rhythm in a relative manner with respect to the salient periodicities of a music excerpt, a cepstroid invariant neural network is proposed to estimate  ...  More recent works incorporated Neural Networks (NN) for handling the Beat-Tracking task.  ... 
doi:10.5281/zenodo.1417736 fatcat:f2cqtllbgzhhxnkfzcw5lp6v2m