Ultra-Low Bitrate Perceptual Image Compression with Shallow Encoder

2025-12-13

AI Summary

To address the deployment challenges of ultra-low-bitrate (<0.05 bpp) image compression on resource-constrained edge devices, this paper proposes AEIC—an asymmetric, minimalist compression framework. AEIC employs a lightweight, shallow CNN encoder coupled with a single-step diffusion-based decoder, and introduces a novel bidirectional feature distillation mechanism to enhance low-dimensional representation capability. Furthermore, it adopts a perception-driven rate-distortion optimization strategy during training. The encoder achieves extreme efficiency—35.8 FPS for 1080p images—while the decoder significantly improves reconstruction quality and perceptual fidelity. Extensive experiments demonstrate that AEIC consistently outperforms state-of-the-art methods across the joint rate–distortion–perception trade-off, establishing a new, efficient, and practical paradigm for ultra-low-bitrate compression in compute-limited scenarios.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Ultra-low bitrate image compression (below 0.05 bits per pixel) is increasingly critical for bandwidth-constrained and computation-limited encoding scenarios such as edge devices. Existing frameworks typically rely on large pretrained encoders (e.g., VAEs or tokenizer-based models) and perform transform coding within their generative latent space. While these approaches achieve impressive perceptual fidelity, their reliance on heavy encoder networks makes them unsuitable for deployment on weak sender devices. In this work, we explore the feasibility of applying shallow encoders for ultra-low bitrate compression and propose a novel Asymmetric Extreme Image Compression (AEIC) framework that pursues simultaneously encoding simplicity and decoding quality. Specifically, AEIC employs moderate or even shallow encoder networks, while leveraging an one-step diffusion decoder to maintain high-fidelity and high-realism reconstructions under extreme bitrates. To further enhance the efficiency of shallow encoders, we design a dual-side feature distillation scheme that transfers knowledge from AEIC with moderate encoders to its shallow encoder variants. Experiments demonstrate that AEIC not only outperforms existing methods on rate-distortion-perception performance at ultra-low bitrates, but also delivers exceptional encoding efficiency for 35.8 FPS on 1080P input images, while maintaining competitive decoding speed compared to existing methods.

Problem

• Develops shallow encoder for ultra-low bitrate image compression
• Proposes asymmetric framework to balance encoding simplicity and decoding quality
• Enhances efficiency via dual-side feature distillation for edge devices

Innovation

• Shallow encoder networks for ultra-low bitrate compression
• One-step diffusion decoder for high-fidelity reconstruction
• Dual-side feature distillation to enhance shallow encoder efficiency