SELF-ORGANIZED VARIATIONAL AUTOENCODERS (SELF-VAE) FOR LEARNED IMAGE COMPRESSION
Author | Yílmaz, M. Akín |
Author | Kelesş, Onur |
Author | Güven, Hilal |
Author | Tekalp, A. Murat |
Author | Malik, Junaid |
Author | Kíranyaz, Serkan |
Available date | 2023-09-24T08:57:18Z |
Publication Date | 2021 |
Publication Name | Proceedings - International Conference on Image Processing, ICIP |
Resource | Scopus |
ISSN | 2381-8549 |
Abstract | In end-to-end optimized learned image compression, it is standard practice to use a convolutional variational autoencoder with generalized divisive normalization (GDN) to transform images into a latent space. Recently, Operational Neural Networks (ONNs) that learn the best non-linearity from a set of alternatives, and their "self-organized" variants, Self-ONNs, that approximate any non-linearity via Taylor series have been proposed to address the limitations of convolutional layers and a fixed nonlinear activation. In this paper, we propose to replace the convolutional and GDN layers in the variational autoencoder with self-organized operational layers, and propose a novel self-organized variational autoencoder (Self-VAE) architecture that benefits from stronger non-linearity. The experimental results demonstrate that the proposed Self-VAE yields improvements in both rate-distortion performance and perceptual image quality. |
Sponsor | This work was supported by TUBITAK projects 217E033 and 120C156, and a grant from Turkish Is Bank to KUIS AI Center. A. M. Tekalp also acknowledges support from Turkish Academy of Sciences (TUBA). |
Language | en |
Publisher | IEEE Computer Society |
Subject | End-to-end learned image compression Perceptual quality metrics Rate-distortion performance Self-organized operational layer Variational autoencoder |
Type | Conference Paper |
Pagination | 3732-3736 |
Volume Number | 2021-September |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |
This item appears in the following Collection(s)
-
Electrical Engineering [2611 items ]