Track trials, findings, and clinical evidence as it lands.
Search papers, labs, and topics across Lattice.
To address the tension between predictor capacity and parameter transmission overhead in coordinate-based overfitted codecs, the authors develop LC3EM, an entropy model that augments a lightweight MLP with a parameter-free neighborhood-based linear extrapolation mode. This dual-mode formulation captures long-range directional context via soft, entropy-minimized continuous gating while utilizing an iterative latent rounding stage to resolve continuous-to-discrete quantization discrepancies. Benchmarked on top of COOL-CHIC 4.0 and 5.0 backbones, the method consistently improves compression efficiency, achieving BD-rate reductions of up to 7.69% on structured and computer-generated imagery.
Overfitted neural codecs do not need heavier networks to capture long-range context: pairing a tiny MLP with parameter-free linear extrapolation slashes BD-rates by up to 7.69% with virtually zero parameter transmission overhead.
Coordinate-based overfitting image codecs have attracted increasing attention for their low decoding complexity and independence from cross-image generalization. However, representative approaches such as COOL-CHIC face an inherent entropy-modeling trade-off: lightweight models have limited capacity, while more expressive ones incur additional bitrate overhead from transmitting image-specific parameters. Inspired by the prediction mechanism in traditional codecs, we propose a new entropy-modeling strategy that introduces complementary prediction modes with region-adaptive soft mode selection, rather than relying on a single learned predictor to model diverse types of redundancy. Based on this concept, we develop a Long-Range Context Extrapolation Enhanced Entropy Model (LC3EM), which can be integrated into coordinate-based overfitting codecs. Specifically, a parameter-free Neighborhood-based Linear Extrapolation Mode (NLEM) complements the tiny MLP-based local predictor to exploit long-range contextual redundancy and strongly directional structures. A Minimum-Entropy-Inspired Continuous Mode Selection strategy is designed to adaptively fuse these two complementary modes, while requiring the transmission of only the parameters of a single additional linear layer. Moreover, to alleviate the mismatch between training-time relaxed and actual discrete quantization, we introduce a lightweight iterative latent rounding refinement stage to improve compression performance. Experiments demonstrate consistent improvements across diverse benchmarks, particularly on highly regular computer-generated images. When integrated with COOL-CHIC 4.0, the proposed method achieves BD-rate gains of -3.43\% and -7.69\% on the SIQAD and API datasets, respectively. With COOL-CHIC 5.0 as the backbone, the corresponding gains are -2.88\% and -3.15\%, respectively. The code will be made publicly available soon.