ROI-based Deep Image Compression with Implicit Bit Allocation

2025-11-12

AI Summary

In ROI image compression, explicit bit allocation disrupts the statistical distribution of entropy models, degrading coding efficiency. To address this, we propose an implicit bit allocation framework that replaces conventional hard masking of background regions with a mask-guided feature enhancement module—incorporating region-adaptive attention and frequency-spatial collaborative attention—to enable soft, region-aware bit resource allocation. Additionally, a dual-decoder architecture is introduced to separately optimize foreground enhancement and background fidelity reconstruction. This work is the first to integrate implicit bit allocation with frequency-spatial collaborative modeling, preserving entropy model statistical stability while improving rate-distortion performance. Evaluated on COCO2017, our method significantly outperforms existing explicit allocation approaches, achieving superior visual quality and background reconstruction fidelity—especially under high compression ratios.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Region of Interest (ROI)-based image compression has rapidly developed due to its ability to maintain high fidelity in important regions while reducing data redundancy. However, existing compression methods primarily apply masks to suppress background information before quantization. This explicit bit allocation strategy, which uses hard gating, significantly impacts the statistical distribution of the entropy model, thereby limiting the coding performance of the compression model. In response, this work proposes an efficient ROI-based deep image compression model with implicit bit allocation. To better utilize ROI masks for implicit bit allocation, this paper proposes a novel Mask-Guided Feature Enhancement (MGFE) module, comprising a Region-Adaptive Attention (RAA) block and a Frequency-Spatial Collaborative Attention (FSCA) block. This module allows for flexible bit allocation across different regions while enhancing global and local features through frequencyspatial domain collaboration. Additionally, we use dual decoders to separately reconstruct foreground and background images, enabling the coding network to optimally balance foreground enhancement and background quality preservation in a datadriven manner. To the best of our knowledge, this is the first work to utilize implicit bit allocation for high-quality regionadaptive coding. Experiments on the COCO2017 dataset show that our implicit-based image compression method significantly outperforms explicit bit allocation approaches in rate-distortion performance, achieving optimal results while maintaining satisfactory visual quality in the reconstructed background regions.

Problem

• Improving ROI-based image compression with implicit bit allocation strategy
• Enhancing coding performance by preserving entropy model statistical distribution
• Balancing foreground enhancement and background quality preservation adaptively

Innovation

• Implicit bit allocation replaces explicit masking
• Mask-guided feature enhancement module with dual attention
• Dual decoders separately reconstruct foreground and background