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

Energy Losses in Transformers01:21

Energy Losses in Transformers

In an ideal transformer, it is assumed that there are no energy losses, and, hence, all the power at the primary winding is transferred to the secondary winding. However, in reality,  the transformers always have some energy losses, and, hence, the output power obtained at the secondary winding is less than the input power at the primary winding due to energy losses.
There are four main reasons for energy losses in transformers.
The first cause can be  the high resistance of the copper windings...
Reducing Line Loss01:18

Reducing Line Loss

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 in...
Downsampling01:20

Downsampling

When considering a sampled sequence with zero values between sampling instants, one can replace it by taking every N-th value of the sequence. At these integer multiples of N, the original and sampled sequences coincide. This process, known as decimation, involves extracting every N-th sample from a sequence, thereby creating a more efficient sequence.
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
Lossy Lines and Overvoltages01:22

Lossy Lines and Overvoltages

Transmission-line series resistance and shunt conductance cause three primary effects: attenuation, distortion, and power losses.
Attenuation
When constant series resistance and shunt conductance are present, voltage and current equations are modified. The propagation constant indicates that voltage and current waves consist of both forward and backward traveling components. These waves attenuate as they propagate, with the attenuation factor related to the resistance and conductance. In a...
Entropy and the Second Law of Thermodynamics01:20

Entropy and the Second Law of Thermodynamics

The second law of thermodynamics can be stated quantitatively using the concept of entropy. Entropy is the measure of disorder of the system.
The relation  between entropy and disorder can be illustrated with the example of the phase change of ice to water. In ice, the molecules are located at specific sites giving a solid state, whereas, in a liquid form, these molecules are much freer to move. The molecular arrangement has therefore become more randomized. Although the change in average...
Entropy and the Second Law of Thermodynamics01:26

Entropy and the Second Law of Thermodynamics

Consider an isolated system in which a hot object is placed in contact with a cold one. This is an irreversible process that eventually leads both objects to reach the same equilibrium temperature. It is crucial to note that the constituents of any substance exhibit increased disorder at higher temperatures. As a cold substance absorbs heat, its constituents become more disordered. The energy transfer from a hotter object to a cooler one increases the system's disorder or randomness. This...

You might also read

Related Articles

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

Sort by
Same author

Trainable clustering framework for spatial transcriptomics.

Bioinformatics advances·2026
Same author

A deep reinforcement based echo state network for network intrusion classification.

PloS one·2026
Same author

The good, the bad, and the ugly: opportunities, challenges, and pitfalls in spatial proteomics modeling.

Briefings in bioinformatics·2026
Same author

Spatial information matters: are traditional imputation methods effective for spatial transcriptomics data?

Briefings in bioinformatics·2026
Same author

Comprehensive retail network dataset for bangladesh: Shop-level sales and location data.

Data in brief·2026
Same author

Adaptive TreeHive: Ensemble of trees for enhancing imbalanced intrusion classification.

PloS one·2025
Same journal

Exploiting audio-visual modalities in videos: Object detection via multi-stage bilateral coupling network.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Reliability-aware modality completion with cross-modal distillation for federated learning with missing modalities.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

IGFD-Net: Illumination-guided frequency decoupling for polarization image fusion.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Multiple-Strategies dung beetle optimizer and its applications in engineering optimization and bankruptcy prediction.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Aggregating global-scale pixel-wise forgery cues within a graph.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Finite-Time intermittent control for secure synchronization of Neutral-Type stochastic delayed neural networks under aperiodic DoS attacks.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

Related Experiment Videos

TextEconomizer: Enhancing lossy text compression with denoising transformers and entropy coding.

Mahbub E Sobhani1, Anika Tasnim Rodela1, Chowdhury Mofizur Rahman2

  • 1United International University, Department of Computer Science and Engineering, Dhaka, Bangladesh.

Neural Networks : the Official Journal of the International Neural Network Society
|May 26, 2026
PubMed
Summary
This summary is machine-generated.

TextEconomizer achieves significant data size reduction for text using lossy compression, maintaining core meaning with high fidelity. This novel framework offers superior space efficiency and fewer parameters than existing models.

Keywords:
Auto-encoderDecoderDenoising transformerEncoderText compressionTextEconomizerTransformer

Related Experiment Videos

Area of Science:

  • Natural Language Processing
  • Machine Learning
  • Data Compression

Background:

  • Lossy text compression is crucial for applications like summarization and digital archives where exact fidelity is not paramount.
  • Existing transformer models dominate language tasks but lack integrated context vectors and entropy coding for efficient text generation.
  • Previous research focused on near-lossless token generation, often neglecting space efficiency in text compression.

Purpose of the Study:

  • To introduce TextEconomizer, a novel encoder-decoder framework for efficient lossy text compression.
  • To explore the integration of context vectors and entropy coding within transformer networks for enhanced storage efficiency.
  • To achieve high-quality text outputs with significant size reduction and minimal parameter count.

Main Methods:

  • Developed TextEconomizer, a transformer-based encoder-decoder framework utilizing latent representations for input reduction.
  • Incorporated entropy coding into the transformer architecture to improve storage efficiency.
  • Evaluated performance using metrics like BLEU, ROUGE, METEOR, and semantic similarity, and compared against LSTM autoencoders and a modified transformer (LLaMAFormer).

Main Results:

  • TextEconomizer reduced variable-sized inputs by 50% to 80% without prior dataset knowledge, achieving compression ratios up to 5.39×.
  • The framework demonstrated near-perfect text quality and operated with approximately 153 times fewer parameters than comparable models.
  • An LSTM autoencoder achieved a 67× compression ratio with 196 times fewer parameters; LLaMAFormer reduced parameters by 263-fold compared to ICAE while maintaining quality.

Conclusions:

  • TextEconomizer offers a significant breakthrough in lossy text compression, effectively balancing memory efficiency and high-fidelity outputs.
  • The framework surpasses existing transformer-based models in space utilization and parameter efficiency.
  • This approach enables optimal space utilization for large text datasets, advancing the field of efficient natural language processing.