Reducing image transmission size by using lower-quality versions opportunistically
Opportunistic Conditional Entropy Coding with Frozen Analysis and Synthesis Transforms
Information Theory
Summary
When you want a better-quality image, sometimes your device already has a lower-quality version of it downloaded. Traditional methods don’t take advantage of this when sending a better image next. The authors developed a way for image compressors to use that lower-quality image if it’s there to reduce the amount of data sent. If the lower-quality image is missing, their method still works normally without much penalty. This helps send improved images faster and with less data.
What this means in practice
- •For video streaming services: Decrease bandwidth use for streaming higher-quality images when users may already have lower-quality versions stored locally.
- •For mobile app developers: Optimize image updates by sending only incremental data when users already have a prior version of the image.
Authors
Vincent Corlay, Maxime Rousselot, Andriy Enttsel
Abstract
In many delivery settings, a receiver may already hold a lower-quality or lower-resolution representation of an image, obtained through an independent transmission. Conventional codecs encode a subsequently requested higher-quality representation without exploiting this incidental side information, whereas conditional codecs generally assume a prescribed source of side information that is always available. We instead consider an opportunistic setting in which side information may or may not be present. We introduce a single entropy model that conditions on a previously decoded latent when available and falls back to a standard hyperprior otherwise. The proposed adapter maps the side-information latent to the prior signal required by the entropy model, allowing the same model to support multiple target and side-information quality combinations. The analysis and synthesis transforms remain frozen, enabling retrofitting of an existing learned codec while preserving its latent representation and reconstruction path. When the receiver holds the quality immediately below the target, the proposed method reduces the rate of the subsequent transmission by up to 46%, or by 52% when an additional hyper-latent is transmitted. In the absence of side information, the rate penalty remains below 4%, and the reconstructions are bit-identical across the conditional and fallback modes.