Don’t Look Back: An Online Beat Tracking Method Using RNN and Enhanced Particle Filtering

Heydari, Mojtaba; Duan, Zhiyao

doi:10.1109/icassp39728.2021.9413915

Cited by 11 publications

(14 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be inferred according to the key equations below. For more detailed information, please refer to our previous work [31].…”

Section: Particle Filtering Inferencementioning

confidence: 99%

“…Aubio [9] 57.09 -BeatNet 75.44 46.49 Böck ACF [4] 64.63 -Böck FF [6,20] 74.18 -DLB [31] 73.77 -IBT [11] 68.99 -…”

Section: Gtzan Datasetmentioning

confidence: 99%

“…However, most music rhythmic analysis approaches that utilize particle filtering, e.g., [27][28][29][30], are classical and do not incorporate neural networks. Alternatively, in our previous work [31], we utilized a particle filtering inference model to infer beat positions using the activations produced by an RNN in an online fashion, but that approach does not attempt to estimate downbeats nor meter.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

BeatNet: CRNN and Particle Filtering for Online Joint Beat Downbeat and Meter Tracking

Heydari¹,

Cwitkowitz²,

Duan³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

The online estimation of rhythmic information, such as beat positions, downbeat positions, and meter, is critical for many real-time music applications. Musical rhythm comprises complex hierarchical relationships across time, rendering its analysis intrinsically challenging and at times subjective. Furthermore, systems which attempt to estimate rhythmic information in real-time must be causal and must produce estimates quickly and efficiently. In this work, we introduce an online system for joint beat, downbeat, and meter tracking, which utilizes causal convolutional and recurrent layers, followed by a pair of sequential Monte Carlo particle filters applied during inference. The proposed system does not need to be primed with a time signature in order to perform downbeat tracking, and is instead able to estimate meter and adjust the predictions over time. Additionally, we propose an information gate strategy to significantly decrease the computational cost of particle filtering during the inference step, making the system much faster than previous sampling-based methods. Experiments on the GTZAN dataset, which is unseen during training, show that the system outperforms various online beat and downbeat tracking systems and achieves comparable performance to a baseline offline joint method.

show abstract

“…It can be inferred according to the key equations below. For more detailed information, please refer to our previous work [31].…”

Section: Particle Filtering Inferencementioning

confidence: 99%

“…Aubio [9] 57.09 -BeatNet 75.44 46.49 Böck ACF [4] 64.63 -Böck FF [6,20] 74.18 -DLB [31] 73.77 -IBT [11] 68.99 -…”

Section: Gtzan Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

BeatNet: CRNN and Particle Filtering for Online Joint Beat Downbeat and Meter Tracking

Heydari¹,

Cwitkowitz²,

Duan³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Such approaches can be split into two groups: unrolling-based methods and prior learning methods. Unrolling-based approaches integrate physical models into the learning process by unfolding each iteration of the classical optimization problem as a layer of a neural network which includes algorithms such as PINN [13], PI-GAN [14], [15] and [16]. These approaches provide improved accuracy while they are time-consuming as they require network retraining for each iteration.…”

Section: Introductionmentioning

confidence: 99%

Regularization by Adversarial Learning for Ultrasound Elasticity Imaging

Mohammadi¹,

Doyley²,

Çetin³

2021

Preprint

View full text Add to dashboard Cite

Classical model-based imaging methods for ultrasound elasticity inverse problem require prior constraints about the underlying elasticity patterns, while finding the appropriate hand-crafted prior for each tissue type is a challenge. In contrast, standard data-driven methods count solely on supervised learning on the training data pairs leading to massive network parameters for unnecessary physical model relearning which might not be consistent with the governing physical models of the imaging system. Fusing the physical forward model and noise statistics with data-adaptive priors leads to a united reconstruction framework that guarantees the learned reconstruction agrees with the physical models while coping with the limited training data. In this paper, we propose a new methodology for estimating the elasticity image by solving a regularized optimization problem which benefits from the physics-based modeling via a data-fidelity term and adversarially learned priors via a regularization term. In this method, the regularizer is trained based on the Wasserstein Generative Adversarial Network (WGAN) objective function which tries to distinguish the distribution of clean and noisy images. Leveraging such an adversarial regularizer for parameterizing the distribution of latent images and using gradient descent (GD) for solving the corresponding regularized optimization task leads to stability and convergence of the reconstruction compared to pixel-wise supervised learning schemes. Our simulation results verify the effectiveness and robustness of the proposed methodology with limited training datasets.

show abstract

“…Unrolling-based methods infuse the physical imaging model into the learning procedure by unrolling every single iteration of the optimization task as a neural network layer. This type of approaches including PINN [11], PI-GAN [12], [13], [14], [15], and MoDL [16] which perform network retraining at each optimization task iteration. On the other hand, prior learning techniques try to have learned units embedded in model-based image reconstruction by bringing learned priors as data-driven regularizers into physics-model-based image reconstruction.…”

Section: Introductionmentioning

confidence: 99%

MR Elasticity Reconstruction Using Statistical Physical Modeling and Explicit Data-Driven Denoising Regularizer

Mohammadi

Doyley

Çetin

2021

2021 IEEE Statistical Signal Processing Workshop (SSP)

View full text Add to dashboard Cite

Elasticity image, visualizing the quantitative map of tissue stiffness, can be reconstructed by solving an inverse problem. Classical methods for magnetic resonance elastography (MRE) try to solve a regularized optimization problem comprising a deterministic physical model and a prior constraint as data-fidelity term and regularization term, respectively. For improving the elasticity reconstructions, appropriate prior about the underlying elasticity distribution is required which is not unique. This article proposes an infused approach for MRE reconstruction by integrating the statistical representation of the physical laws of harmonic motions and learning-based prior. For data-fidelity term, we use a statistical linear-algebraic model of equilibrium equations and for the regularizer, data-driven regularization by denoising (RED) is utilized. In the proposed optimization paradigm, the regularizer gradient is simply replaced by the residual of learned denoiser leading to time-efficient computation and convex explicit objective function. Simulation results of elasticity reconstruction verify the effectiveness of the proposed approach.

show abstract

Don’t Look Back: An Online Beat Tracking Method Using RNN and Enhanced Particle Filtering

Cited by 11 publications

References 11 publications

BeatNet: CRNN and Particle Filtering for Online Joint Beat Downbeat and Meter Tracking

BeatNet: CRNN and Particle Filtering for Online Joint Beat Downbeat and Meter Tracking

Regularization by Adversarial Learning for Ultrasound Elasticity Imaging

MR Elasticity Reconstruction Using Statistical Physical Modeling and Explicit Data-Driven Denoising Regularizer

Contact Info

Product

Resources

About