Key points are not available for this paper at this time.
Learned image compression codecs have recently achieved impressive compression performances surpassing the most efficient image coding architectures. However, most approaches are trained to minimize rate and distortion which often leads to unsatisfactory visual results at low bitrates since perceptual metrics are not taken into account. In this paper, we show that conditional diffusion models can lead to promising results in the generative compression task when used as a decoder, and that, given a compressed representation, they allow creating new tradeoff points between distortion and perception at the decoder side based on the sampling method.
Building similarity graph...
Analyzing shared references across papers
Loading...
Mari et al. (Tue,) studied this question.
synapsesocial.com/papers/68e75b28b6db6435876d247b — DOI: https://doi.org/10.48550/arxiv.2403.02887
Daniele Mari
University of Padua
Simone Milani
University of Padua
Building similarity graph...
Analyzing shared references across papers
Loading...