Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Reducing Line Loss01:18

Reducing Line Loss

227
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
227
Improving Translational Accuracy02:07

Improving Translational Accuracy

12.1K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
12.1K
Improving Translational Accuracy02:07

Improving Translational Accuracy

3.1K
3.1K
Linear Approximation in Frequency Domain01:26

Linear Approximation in Frequency Domain

204
Linear systems are characterized by two main properties: superposition and homogeneity. Superposition allows the response to multiple inputs to be the sum of the responses to each individual input. Homogeneity ensures that scaling an input by a scalar results in the response being scaled by the same scalar.
In contrast, nonlinear systems do not inherently possess these properties. However, for small deviations around an operating point, a nonlinear system can often be approximated as linear....
204
Impression Management Techniques IV: Altercasting01:14

Impression Management Techniques IV: Altercasting

38
Altercasting is a strategic communication technique in which an individual imposes a specific identity or social role onto another person to influence their behavior and shape the interaction. By presuming a role—such as “responsible leader” or “patient person”—altercasting encourages the target to conform to that identity, often aligning their behavior with the expectations associated with the role. The power of this tactic lies in its subtlety; once a role...
38
Reconstruction of Signal using Interpolation01:10

Reconstruction of Signal using Interpolation

430
Signal processing techniques are essential for accurately converting continuous signals to digital formats and vice versa. When a continuous signal is sampled with a period T, the resulting sampled signal exhibits replicas of the original spectrum in the frequency domain, spaced at intervals equal to the sampling frequency. To handle this sampled signal, a zero-order hold method can be applied, which creates a piecewise constant signal by retaining each sample's value until the next...
430

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A speech prediction model based on codec modeling and transformer decoding.

Computer speech & language·2026
Same author

A Molecular Trimming Strategy for Hypoxia-Tolerant Photosensitizers With Enhanced cGAS-STING Activation.

Angewandte Chemie (International ed. in English)·2026
Same author

Towards decoupling frontend enhancement and backend recognition in monaural robust ASR.

Computer speech & language·2026
Same author

Peripheral mechanisms of tactile sensation in fish.

Current opinion in neurobiology·2026
Same author

Efficacy of SWIM technology combined with direct aspiration first pass technique for large vessel occlusion in acute ischemic stroke.

American journal of translational research·2026
Same author

Pharmacological activation of HSF1 by HSF1A mitigates heatstroke-induced acute kidney injury via ferroptosis inhibition.

International journal of biological macromolecules·2026
Same journal

<math></math> Estimation and Voicing Detection With Cascade Architecture in Noisy Speech.

IEEE/ACM transactions on audio, speech, and language processing·2025
Same journal

Speech Enhancement for Cochlear Implant Recipients using Deep Complex Convolution Transformer with Frequency Transformation.

IEEE/ACM transactions on audio, speech, and language processing·2025
Same journal

Selective Acoustic Feature Enhancement for Speech Emotion Recognition With Noisy Speech.

IEEE/ACM transactions on audio, speech, and language processing·2024
Same journal

Glottal Airflow Estimation using Neck Surface Acceleration and Low-Order Kalman Smoothing.

IEEE/ACM transactions on audio, speech, and language processing·2023
Same journal

Bilateral Cochlear Implant Processing of Coding Strategies With CCi-MOBILE, an Open-Source Research Platform.

IEEE/ACM transactions on audio, speech, and language processing·2023
Same journal

Robust Vocal Quality Feature Embeddings for Dysphonic Voice Detection.

IEEE/ACM transactions on audio, speech, and language processing·2023
See all related articles

Related Experiment Video

Updated: Oct 31, 2025

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.7K

Towards Model Compression for Deep Learning Based Speech Enhancement.

Ke Tan1, DeLiang Wang2

  • 1Department of Computer Science and Engineering, The Ohio State University, Columbus, OH, 43210-1277 USA.

IEEE/ACM Transactions on Audio, Speech, and Language Processing
|June 28, 2021
PubMed
Summary
This summary is machine-generated.

Deep neural network (DNN) models for speech enhancement can be compressed using sparse regularization, iterative pruning, and quantization. This approach significantly reduces model size without sacrificing performance, enabling deployment on resource-constrained devices.

Keywords:
Model compressionpruningquantizationsparse regularizationspeech enhancement

More Related Videos

Author Spotlight: Advancements in the Fabrication of Synthetic Vocal Fold Models for Phonetic and Robotic Applications
06:24

Author Spotlight: Advancements in the Fabrication of Synthetic Vocal Fold Models for Phonetic and Robotic Applications

Published on: January 5, 2024

1.1K
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

802

Related Experiment Videos

Last Updated: Oct 31, 2025

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.7K
Author Spotlight: Advancements in the Fabrication of Synthetic Vocal Fold Models for Phonetic and Robotic Applications
06:24

Author Spotlight: Advancements in the Fabrication of Synthetic Vocal Fold Models for Phonetic and Robotic Applications

Published on: January 5, 2024

1.1K
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

802

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Signal Processing

Background:

  • Deep neural networks (DNNs) have significantly improved speech enhancement.
  • Large DNN models are computationally intensive and memory-consuming, hindering deployment on edge devices.
  • Existing speech enhancement methods face challenges with resource limitations and latency requirements.

Purpose of the Study:

  • To propose and evaluate compression pipelines for DNN-based speech enhancement.
  • To reduce the model size of speech enhancement systems.
  • To enable efficient deployment of speech enhancement on devices with limited resources.

Main Methods:

  • Developed two compression pipelines for DNN-based speech enhancement.
  • Incorporated three compression techniques: sparse regularization, iterative pruning, and clustering-based quantization.
  • Systematically investigated and evaluated the effectiveness of these techniques and pipelines.

Main Results:

  • Achieved significant reductions in model size for four different DNN models.
  • Maintained high speech enhancement performance despite substantial model compression.
  • Demonstrated effectiveness on speaker separation tasks, indicating broader applicability.

Conclusions:

  • The proposed compression pipelines effectively reduce DNN model sizes for speech enhancement.
  • The compression methods allow for deployment on hardware with limited resources and strict latency needs.
  • The approach is effective for compressing speech separation models, showcasing versatility.