Overview

Rate-distortion theory asks how many bits are needed to represent a source while keeping reconstruction errors within an acceptable level. Classical bounds guide methods such as scalar, vector, and transform quantization, while learned codecs use neural representations and rate allocation to approach these limits. Current work extends the objective to perceptual realism, semantic meaning, downstream task utility, and resource costs in settings ranging from images and video to agent memory and distributed communication. A central tension is that standard distortion measures may not reflect what people perceive or tasks require, while guarantees for richer objectives often depend on restrictive assumptions.

Latest papers

31 papers4 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

Feb 16, 2026

On the Rate-Distortion-Complexity Tradeoff for Semantic Communication

Semantic communication is a novel communication paradigm that focuses on conveying the user's intended meaning rather than the bit-wise transmission of source signals. One of the key challenges is to effectively represent and extract the semantic meaning of any given source signals. While deep learning (DL)-based solutions have shown promising results in extracting implicit semantic information from a wide range of sources, existing work often overlooks the high computational complexity inherent in both model training and inference for the DL-based encoder and decoder. To bridge this gap, this paper proposes a rate-distortion-complexity (RDC) framework which extends the classical rate-distortion theory by incorporating the constraints on semantic distance, including both the traditional bit-wise distortion metric and statistical difference-based divergence metric, and complexity measure, adopted from the theory of minimum description length and information bottleneck. We derive the closed-form theoretical results of the minimum achievable rate under given constraints on semantic distance and complexity for both Gaussian and binary semantic sources. Our theoretical results show a fundamental three-way tradeoff among achievable rate, semantic distance, and model complexity. Extensive experiments on real-world image and video datasets validate this tradeoff and further demonstrate that our information-theoretic complexity measure effectively correlates with practical computational costs, guiding efficient system design in resource-constrained scenarios.