HAMP-LIC: Hessian-Aware Mixed-Precision Post-Training Quantization for Learned Image Compression
2026-08-12 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial IntelligenceMultimedia
AI summaryⓘ
The authors developed a method called HAMP-LIC to make image compression models smaller and faster without losing much quality. They use a smart way to figure out which parts of the model are more sensitive to being simplified, then adjust how much they compress each part. Their approach helps keep the images looking good while making the model run efficiently on different hardware. Tests showed it works better than other methods and avoids errors when moving between devices.
Learned image compressionPost-training quantizationMixed-precision quantizationHessian traceRate-distortionBit width allocationModel compressionEncoding-decoding mismatchCalibration set
Authors
Yuefeng Zhang
Abstract
Use this plain-text version for the arXiv abstract field: Learned image compression (LIC) models achieve strong rate-distortion performance but are hindered by high computational complexity and encoding-decoding mismatches across heterogeneous hardware platforms. Uniform fixed-precision quantization alleviates these issues but suffers severe quality degradation at low bit widths because it ignores differences in the quantization sensitivities of individual layers. To enable efficient and accurate low-bit deployment of pretrained LIC models, we propose HAMP-LIC, a Hessian-aware mixed-precision post-training quantization (PTQ) framework with a four-stage optimization strategy. First, block-wise sensitivity is estimated from the Hessian trace to capture second-order importance. Second, a task-aware refinement module adjusts these sensitivities by jointly considering quantization distortion and rate-distortion performance. Third, guided by the refined sensitivity profile, bit widths are allocated under a global model-size constraint to balance efficiency and reconstruction quality. Finally, block-wise reconstruction using a small calibration set further suppresses quantization error. Experiments on representative LIC models, including Minnen2018 and Cheng2020, demonstrate that HAMP-LIC achieves up to 4.85x model compression with as little as 0.59% BD-rate loss. It consistently outperforms existing fixed- and mixed-precision PTQ methods across multiple datasets while completely eliminating cross-platform encoding-decoding errors.