Near-lossless coding schemes traditionally rely on uniform quantization to control the maximum absolute error (L ∞ norm) of residual signals, often assuming a parametric model for the source distribution. This paper introduces a novel design framework for non-uniform, entropy-aware L ∞ -oriented scalar quantizers that leverages a tight and differentiable approximation of the L ∞ distortion metric and does not require any parametric density function formulations. The framework is evaluated on both synthetic parametric sources and real-world medical depth map video datasets. For smoothly decaying distributions, such as the continuous Laplacian or discrete two sided geometric distributions, the pro- posed method naturally converges to near uniform quantizers, consistent with theoretical expectations. In contrast, for sparse or irregular sources, the algorithm produces highly non-uniform bin allocations that adapt to the local distribution structure and improve rate-distortion efficiency. When embedded in a residual-based near-lossless compression scheme, the resulting codec consistently outperforms versions equipped with uniform or piecewise-uniform quantizers, as well as state-of-the-art near-lossless schemes such as JPEG-LS and CALIC.
Alecu, A, Tahouri, MA, Munteanu, A & Pavaloiu, B 2025, 'Non-Uniform Entropy-Constrained L ∞ Quantization for Sparse and Irregular Sources', Entropy, vol. 27, no. 11, 1126. https://doi.org/10.3390/e27111126
Alecu, A., Tahouri, M. A., Munteanu, A., & Pavaloiu, B. (2025). Non-Uniform Entropy-Constrained L ∞ Quantization for Sparse and Irregular Sources. Entropy, 27(11), Article 1126. https://doi.org/10.3390/e27111126
@article{9899520eae3d414b80349a4bab03c362,
title = "Non-Uniform Entropy-Constrained L ∞ Quantization for Sparse and Irregular Sources",
abstract = "Near-lossless coding schemes traditionally rely on uniform quantization to control the maximum absolute error (L ∞ norm) of residual signals, often assuming a parametric model for the source distribution. This paper introduces a novel design framework for non-uniform, entropy-aware L ∞ -oriented scalar quantizers that leverages a tight and differentiable approximation of the L ∞ distortion metric and does not require any parametric density function formulations. The framework is evaluated on both synthetic parametric sources and real-world medical depth map video datasets. For smoothly decaying distributions, such as the continuous Laplacian or discrete two sided geometric distributions, the pro- posed method naturally converges to near uniform quantizers, consistent with theoretical expectations. In contrast, for sparse or irregular sources, the algorithm produces highly non-uniform bin allocations that adapt to the local distribution structure and improve rate-distortion efficiency. When embedded in a residual-based near-lossless compression scheme, the resulting codec consistently outperforms versions equipped with uniform or piecewise-uniform quantizers, as well as state-of-the-art near-lossless schemes such as JPEG-LS and CALIC.",
author = "Alin Alecu and Tahouri, \{Mohammad Ali\} and Adrian Munteanu and Bujor Pavaloiu",
note = "Publisher Copyright: {\textcopyright} 2025 by the authors.",
year = "2025",
month = oct,
day = "31",
doi = "10.3390/e27111126",
language = "English",
volume = "27",
journal = "Entropy",
issn = "1099-4300",
publisher = "MDPI AG",
number = "11",
}