What does this research mean for the field?

A data-driven procedure based on convolutional variational autoencoders can accurately identify signal pulses in noisy time-series by isolating background noise within a specific region of the latent space. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To develop a data-driven method using convolutional variational autoencoders to identify signal pulses within long, noisy time-series data.

August 23, 2026ParticlesOpen Access

Exploiting the Latent Space of Deep AutoEncoders for the Identification of Signal Pulses in Noisy Time-Series

View Full Paper

Ask AI

Bookmark

Authors

GAGioacchino Alex Anastasi SASebastiano Francesco Albergo MNMarzio De Napoli

Discussion

Member takes

Overview

Simulation study demonstrates accurate signal pulse detection in noisy waveforms using autoencoders, highlighting utility for liquid argon dark matter detectors.

Key Points

To develop a data-driven method using convolutional variational autoencoders to identify signal pulses within long, noisy time-series data.
Trained a convolutional variational autoencoder for 150 epochs on 7,500 synthetic waveforms of 10,000 samples containing non-Gaussian noise and variable-intensity log-normal signals.
Mapped waveforms into a compressed latent space representation to separate background noise from signal-containing waveforms.
Evaluated detection accuracy on a freshly generated test dataset of synthetic waveforms.
Background-only noise waveforms clustered into a distinct, identifiable region within the model's latent space.
Achieved 100% labeling accuracy for test waveforms with signal amplitudes well above baseline noise levels.
Maintained signal detection capability with accuracy decreasing only when signal amplitudes matched accidental noise pulse levels.

Cite This Study

Anastasi et al. (2026) studied this question.

synapsesocial.com/papers/6a8aad417677a3411444560c https://doi.org/10.3390/particles9030085

View Full Paper

Ask AI

Bookmark

August 23, 2026ParticlesOpen Access

Exploiting the Latent Space of Deep AutoEncoders for the Identification of Signal Pulses in Noisy Time-Series

View Full Paper

Ask AI

Bookmark

Authors

GAGioacchino Alex Anastasi SASebastiano Francesco Albergo MNMarzio De Napoli

Discussion

Member takes

Overview

Simulation study demonstrates accurate signal pulse detection in noisy waveforms using autoencoders, highlighting utility for liquid argon dark matter detectors.

Key Points

To develop a data-driven method using convolutional variational autoencoders to identify signal pulses within long, noisy time-series data.
Trained a convolutional variational autoencoder for 150 epochs on 7,500 synthetic waveforms of 10,000 samples containing non-Gaussian noise and variable-intensity log-normal signals.
Mapped waveforms into a compressed latent space representation to separate background noise from signal-containing waveforms.
Evaluated detection accuracy on a freshly generated test dataset of synthetic waveforms.
Background-only noise waveforms clustered into a distinct, identifiable region within the model's latent space.
Achieved 100% labeling accuracy for test waveforms with signal amplitudes well above baseline noise levels.
Maintained signal detection capability with accuracy decreasing only when signal amplitudes matched accidental noise pulse levels.

Cite This Study

Anastasi et al. (2026) studied this question.

synapsesocial.com/papers/6a8aad417677a3411444560c https://doi.org/10.3390/particles9030085

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark