DynaQuant: Dynamic Mixed-Precision Quantization for Learned Image Compression

TL;DR

DynaQuant introduces a dynamic mixed-precision quantization framework for learned image compression, adaptively optimizing bit-width and quantization parameters per layer and input to improve efficiency without sacrificing performance.

Contribution

It presents a novel end-to-end learnable system with content-aware quantization and dynamic bit-width selection, enhancing LIC efficiency and adaptability.

Findings

Achieves comparable rate-distortion performance to full-precision models.

Reduces computational and storage costs significantly.

Enables practical deployment of LIC on diverse hardware.

Abstract

Prevailing quantization techniques in Learned Image Compression (LIC) typically employ a static, uniform bit-width across all layers, failing to adapt to the highly diverse data distributions and sensitivity characteristics inherent in LIC models. This leads to a suboptimal trade-off between performance and efficiency. In this paper, we introduce DynaQuant, a novel framework for dynamic mixed-precision quantization that operates on two complementary levels. First, we propose content-aware quantization, where learnable scaling and offset parameters dynamically adapt to the statistical variations of latent features. This fine-grained adaptation is trained end-to-end using a novel Distance-aware Gradient Modulator (DGM), which provides a more informative learning signal than the standard Straight-Through Estimator. Second, we introduce a data-driven, dynamic bit-width selector that learns to assign an optimal bit precision to each layer, dynamically reconfiguring the network's precision profile based on the input data. Our fully dynamic approach offers substantial flexibility in balancing rate-distortion (R-D) performance and computational cost. Experiments demonstrate that DynaQuant achieves rd performance comparable to full-precision models while significantly reducing computational and storage requirements, thereby enabling the practical deployment of advanced LIC on diverse hardware platforms.