Open Access   Article Go Back

Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques

D.S. Ene1 , V.I.E. Anireh2 , D. Matthias3 , E.O. Bennett4

  1. Information Technology Centre, Rivers State University, Port Harcourt, Nigeria.
  2. Dept. of Computer Science, Rivers State University, Port Harcourt, Nigeria.
  3. Dept. of Computer Science, Rivers State University, Port Harcourt, Nigeria.
  4. Dept. of Computer Science, Rivers State University, Port Harcourt, Nigeria.

Section:Research Paper, Product Type: Journal Paper
Volume-12 , Issue-5 , Page no. 42-53, May-2024

CrossRef-DOI:   https://doi.org/10.26438/ijcse/v12i5.4253

Online published on May 31, 2024

Copyright © D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

View this paper at   Google Scholar | DPI Digital Library

XML View
PDF Download

How to Cite this Paper

  • IEEE Citation
  • MLA Citation
  • APA Citation
  • BibTex Citation
  • RIS Citation

IEEE Style Citation: D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett, “Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques,” International Journal of Computer Sciences and Engineering, Vol.12, Issue.5, pp.42-53, 2024.

MLA Style Citation: D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett "Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques." International Journal of Computer Sciences and Engineering 12.5 (2024): 42-53.

APA Style Citation: D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett, (2024). Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques. International Journal of Computer Sciences and Engineering, 12(5), 42-53.

BibTex Style Citation:
@article{Ene_2024,
author = {D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett},
title = {Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques},
journal = {International Journal of Computer Sciences and Engineering},
issue_date = {5 2024},
volume = {12},
Issue = {5},
month = {5},
year = {2024},
issn = {2347-2693},
pages = {42-53},
url = {https://www.ijcseonline.org/full_paper_view.php?paper_id=5690},
doi = {https://doi.org/10.26438/ijcse/v12i5.4253}
publisher = {IJCSE, Indore, INDIA},
}

RIS Style Citation:
TY - JOUR
DO = {https://doi.org/10.26438/ijcse/v12i5.4253}
UR - https://www.ijcseonline.org/full_paper_view.php?paper_id=5690
TI - Optimization of Variational Autoencoder Model for Mobile Computing Environment Using Amortized Stochastic Variational Inference and Miniaturization Techniques
T2 - International Journal of Computer Sciences and Engineering
AU - D.S. Ene, V.I.E. Anireh, D. Matthias, E.O. Bennett
PY - 2024
DA - 2024/05/31
PB - IJCSE, Indore, INDIA
SP - 42-53
IS - 5
VL - 12
SN - 2347-2693
ER -

VIEWS PDF XML
137 121 downloads 77 downloads
  
  
           

Abstract

Variational Autoencoders (VAEs) are powerful machine learning models that can be deployed on mobile devices. However, VAEs are often deployed on resource-constrained mobile platforms, resulting in a high computational overhead. In this study, we present a novel framework, called the Miniaturizing Variations Auto Encoder (mVAE), to overcome the computational constraints associated with VAE deployment on mobile platforms. By leveraging advanced miniaturization techniques and integrating Amortized Stochastic Variational Inference (ASVI), this framework unlocks the full potential of VAE models in the mobile realm. Through extensive experiments and performance analysis, we aim to demonstrate the feasibility and efficiency of the m VAE framework in enabling the deployment of sophisticated machine learning applications on mobile systems. The findings of this study not only contribute to the advancement of mobile computing but also pave the way for a wide range of practical applications, empowering mobile users with powerful AI capabilities. Overall, this research contributes not only to theoretical foundations but also provides practical insights into implementation, addressing the need for efficient machine learning systems in mobile computing environments.

Key-Words / Index Term

AI, Autoencoder, Amortized, Variational Inference, VAE, ASVI

References

[1] Z. Chang et al., "A survey of recent advances in edge-computing-powered artificial intelligence of things," IEEE Internet of Things Journal, Vol.8, Issue.18, pp.13849-13875, 2021.
[2] S. N. Ajani et al., "Advancements in Computing: Emerging Trends in Computational Science with Next-Generation Computing," International Journal of Intelligent Systems and Applications in Engineering, Vol.12, Issue.7, pp.546-559, 2024.
[3] R. G. Wei et al., "Variations in variational autoencoders-a comparative evaluation," IEEE Access, Vol.8, pp.153651-153670, 2020.
[4] M. Agiwal et al., "Next generation 5G wireless networks: A comprehensive survey," IEEE, Vol.18, Issue.3, pp.1617-1655, 2016.
[5] D. P. Kingma and M. Welling, "An introduction to variational autoencoders," Foundations and Trends® in Machine Learning, Vol.12, Issue.4, pp.307-392, 2019.
[6] G. Roeder et al., "Efficient amortised Bayesian inference for hierarchical and nonlinear dynamical systems," In International Conference on Machine Learning, PMLR, pp.4445-4455, 2019
[7] J. Paul et al., "Resource-aware programming for robotic vision," arXiv preprint arXiv:1405.2908, 2014.
[8] A. Frickenstein et al., "Resource-aware optimization of dnns for embedded applications," In 2019 16th Conference on Computer and Robot Vision (CRV), IEEE, pp.17-24, 2019.