Latent Channel Energy in Learned Image Codecs
Prativadibhayankaram S, Richter T, Fößel S, Kaup A (2024)
Publication Type: Conference contribution
Publication year: 2024
Publisher: SPIE
Book Volume: 13137
Conference Proceedings Title: Proceedings of SPIE - The International Society for Optical Engineering
Event location: San Diego, CA, USA
ISBN: 9781510679344
DOI: 10.1117/12.3033220
Abstract
With the advent of learned image compression, numerous models have been developed. These models make use of non-linear transforms that are learnt during the training process, where an image is transformed into a latent space, quantized and entropy coded. At the decoder, the quantized latent is recovered and transformed back to image space through a synthesis transform. In this work, we attempt to present an analysis of the energy distribution across channels. In our prior works, we demonstrated the features captured by the analysis transform, that can provide insights into the bitrate distribution across channels. Building on that, we extend our findings with quantitative measurements. We consider various learned image codecs that are based on the variational autoencoder framework and compare them with Karhunen Loève Transform (KLT) in terms of energy compaction. We also compare the closeness of the learned transforms to KLT to study the relationship between the design of classical codecs and learned codecs.
Authors with CRIS profile
Involved external institutions
Germany (DE)How to cite
APA:
Prativadibhayankaram, S., Richter, T., Fößel, S., & Kaup, A. (2024). Latent Channel Energy in Learned Image Codecs. In Andrew G. Tescher, Touradj Ebrahimi (Eds.), Proceedings of SPIE - The International Society for Optical Engineering. San Diego, CA, USA: SPIE.
MLA:
Prativadibhayankaram, Srivatsa, et al. "Latent Channel Energy in Learned Image Codecs." Proceedings of the Applications of Digital Image Processing XLVII 2024, San Diego, CA, USA Ed. Andrew G. Tescher, Touradj Ebrahimi, SPIE, 2024.
BibTeX: Download