An Agentic Framework for Quality-Aware and Scenario-Adaptive Image Compression
Authors: Zhenghao Chen, Lebin Zhou, Zicong Chen, Naifu Xue, Jiahao Li, Bin Li, Dong Xu, Yan Lu
Organizations: The University of Newcastle · Santa Clara University · The Communication University of China · Microsoft Research Asia · The University of Hong Kong
Despite decades of progress, image codecs remain difficult to adapt to diverse real-world compression scenarios for two main reasons. First, most image codecs are optimized for a specific objective and require substantial redevelopment when the compression goal changes. Second, they typically assume clean inputs, whereas sensor-captured images often contain degradations that alter rate-quality behavior. To address these challenges, we propose Practical Image Coding Agents (PICA), a multi-agent framework that adapts existing codecs to diverse user requirements and potentially degraded inputs. PICA consists of four collaborative agents: a Perception agent that interprets user intent and analyzes image content, an Improvement agent that jointly considers bitrate and reconstruction quality to guide pre-coding restoration, a Coding agent that refines quality-control settings from global to regional levels, and an After-coding improvement agent that addresses residual degradations and compression-induced artifacts through restoration guided solely by image quality. To enable codec-agnostic regional adaptation, we introduce Agentic ROI Coding (AROIC), which treats the codec as a black box and allows the agent to automatically schedule, execute, evaluate, and reflect ROI coding plans without architectural modifications or codec-specific retraining. We integrate these four agents with a Scenario-Guided Quality-Rate (SGQR) evaluation strategy to enable more effective scenario-adaptive compression across heterogeneous codecs. Comprehensive experiments on clean, noisy, and blurred images across standard, perceptual, and machine-vision compression scenarios demonstrate that PICA consistently enhances the compression performance of both the conventional H.266/VVC codec and the neural Cheng2020 codec, outperforming competing compression methods on the Kodak, CLIC2020, and HEVC datasets.
Figures & tables
Fig. 1 : Overview of the P ractical I mage C ompression A gent ( PICA ) framework, consisting of four collaborative agents: the P erception agent, the I mprovement agent, the C oding agent, and the A fter-coding improvement agent.
Fig. 2 : The Coding Agent workflow comprises two components: Global QC Search and Agentic ROI Coding (AROIC).
Fig. 3 : Quality-rate performance on the (a) clean, (b) noisy, and (c) blurred Kodak benchmarks. Bitrate is measured in bits per pixel (bpps), while reconstruction quality is evaluated against the original clean reference images.
Fig. 4 : Quality-rate performance on the (a) clean, (b) noisy, and (c) blurred CLIC2020 benchmarks. Bitrate is measured in bits per pixel (bpps), while reconstruction quality is evaluated against the original clean reference images.
Fig. 5 : Quality–rate performance on the (a) clean, (b) noisy, and (c) blurred benchmarks derived from HEVC Classes C and D. Bitrate is measured in bits per pixel (bpp), while object detection performance is evaluated on the reconstructed images using mAP[0.50:0.95] (%). The gray dashed lines indicate detection performance on uncompressed clean inputs.
Fig. 6 : Ablation studies of (a)–(b) the effectiveness of individual agents on the blurred Kodak dataset, where P, I, C, and A denote the Perception, Improvement, Coding, and After-coding Improvement Agents, respectively, and (c) PICA with different VLMs as agent backbones on the noisy Kodak dataset.
Fig. 7 : Qualitative comparison of PICA using VVC and Cheng2020 as base codecs against their corresponding baseline codecs.
Image compression is increasingly used before machine vision systems, yet standard codecs are mainly optimized for human perception rather than downstream task accuracy. We propose SCOPE (Semantic COdec Pre-Editing), a conditional image pre-editor that modifies images before compression to improve the bitrate and detection accuracy trade-off. The model is conditioned on the codec family and compression severity, allowing a single trained network to adapt to multiple compression settings without retraining or loading codec-specific weights. Instead of uniformly preserving the whole image, the pre-editor learns to protect task-relevant regions while simplifying less important background content to improve compressibility. The resulting rate-accuracy curves show consistent gains across different codec families and detector architectures, together with strong transferability to unseen learned and wavelet-based compression families.
Silvia Dani, Leonardo Galteri, Marco Bertini
University of Florence, Florence, Italy · Small Pixels, Florence, Italy
Implicit Neural Representations (INRs) have recently emerged as a promising paradigm for image compression, offering a fundamentally different approach from traditional and learned codecs. Nevertheless, INR-based methods for image compression suffer from long encoding times and a consistent performance gap in classic quality metrics such as PSNR. In this work, we explore the potential of purely INR-based compression methods and we propose PaaF (Picture as a Function), a novel INR-based image codec that introduces improved architectural design, adaptive quantization, and an efficient entropy coding scheme. These components are designed to enhance rate-distortion performance while preserving the simplicity and parallelizability of INR-based decoding. Experimental results demonstrate consistent improvements over existing INR-based methods in both quantitative metrics and perceptual quality. These findings highlight the potential of INR-based approaches and contribute to narrowing the gap between functional representations and more established compression paradigms.
Lorenzo Catania, Dario Allegra
Department of Mathematics and Computer Science, University of Catania, Catania, Italy
Recent advances in generative image compression (GIC) have delivered remarkable improvements in perceptual quality. However, many GICs rely on large-scale and rigid models, which severely constrain their utility for flexible transmission and practical deployment in low-bitrate scenarios. To address these issues, we propose Progressive Generative Image Compression (ProGIC), a compact codec built on residual vector quantization (RVQ). In RVQ, a sequence of vector quantizers encodes the residuals stage by stage, each with its own codebook. The resulting codewords sum to a coarse-to-fine reconstruction and a progressive bitstream, enabling previews from partial data. We pair this with a lightweight backbone based on depthwise-separable convolutions and small attention blocks, enabling practical deployment on both GPUs and CPU-only devices. Experimental results show that ProGIC attains comparable compression performance compared with previous methods. It achieves bitrate savings of up to 57.57% on DISTS and 58.83% on LPIPS compared to MS-ILLM on the Kodak dataset. Beyond perceptual quality, ProGIC enables progressive transmission for flexibility, and also delivers over 10 times faster encoding and decoding compared with MS-ILLM on GPUs for efficiency.
Hao Cao, Chengbin Liang, Wenqi Guo +2
Tsinghua University · State Key Laboratory of Space Network and Communications