Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LatentQGAN: A Hybrid QGAN with Classical Convolutional Autoencoder

Alexis Vieloszynski, Soumaya Cherkaoui, Ola Ahmad, Jean-Frédéric Laprade, Oliver Nahman-Lévesque, Abdallah Aaraba, Shengrui Wang

TL;DR

LatentQGAN, a novel quantum model that uses a hybrid quantum-classical GAN coupled with an auto-encoder, is proposed that holds potential for broader application across various practical data generation tasks.

Abstract

Quantum machine learning consists in taking advantage of quantum computations to generate classical data. A potential application of quantum machine learning is to harness the power of quantum computers for generating classical data, a process essential to a multitude of applications such as enriching training datasets, anomaly detection, and risk management in finance. Given the success of Generative Adversarial Networks in classical image generation, the development of its quantum versions has been actively conducted. However, existing implementations on quantum computers often face significant challenges, such as scalability and training convergence issues. To address these issues, we propose LatentQGAN, a novel quantum model that uses a hybrid quantum-classical GAN coupled with an autoencoder. Although it was initially designed for image generation, the LatentQGAN approach holds potential for broader application across various practical data generation tasks. Experimental outcomes on both classical simulators and noisy intermediate scale quantum computers have demonstrated significant performance enhancements over existing quantum methods, alongside a significant reduction in quantum resources overhead.

LatentQGAN: A Hybrid QGAN with Classical Convolutional Autoencoder

TL;DR

LatentQGAN, a novel quantum model that uses a hybrid quantum-classical GAN coupled with an auto-encoder, is proposed that holds potential for broader application across various practical data generation tasks.

Abstract

Quantum machine learning consists in taking advantage of quantum computations to generate classical data. A potential application of quantum machine learning is to harness the power of quantum computers for generating classical data, a process essential to a multitude of applications such as enriching training datasets, anomaly detection, and risk management in finance. Given the success of Generative Adversarial Networks in classical image generation, the development of its quantum versions has been actively conducted. However, existing implementations on quantum computers often face significant challenges, such as scalability and training convergence issues. To address these issues, we propose LatentQGAN, a novel quantum model that uses a hybrid quantum-classical GAN coupled with an autoencoder. Although it was initially designed for image generation, the LatentQGAN approach holds potential for broader application across various practical data generation tasks. Experimental outcomes on both classical simulators and noisy intermediate scale quantum computers have demonstrated significant performance enhancements over existing quantum methods, alongside a significant reduction in quantum resources overhead.
Paper Structure (13 sections, 10 equations, 6 figures, 1 table)

This paper contains 13 sections, 10 equations, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background
  3. Related Works
  4. LatentQGAN
  5. The Framework of LatentQGAN
  6. Quantum Generative Adversarial Networks (QGAN)
  7. Quantum Generator
  8. Classical Discriminator
  9. The Autoencoder
  10. Experimental details and results
  11. Experimental details
  12. Experimental results
  13. Conclusion

Figures (6)

  • Figure 1: LatentQGAN's overall framework.
  • Figure 2: The autoencoder used in LatentQGAN, with the encoder on the top part, and the decoder on the bottom. The "Normalize" process is represented in red at the end of the encoder.
  • Figure 3: FD results over training. The Fig.\ref{['fig9']} shows a visual inaccuracy and decay respectively before 350 iterations and after 700 iterations
  • Figure 4: Images generated by LatentQGAN for class 3 over 1000 training iterations, on simulator. Before 300 iterations, the results are visually inaccurate, and after 700 iterations, the results visually decay.
  • Figure 5: FD evaluated on many models for comparison with LatentQGAN, on simulator
  • ...and 1 more figures