Search papers, labs, and topics across Lattice.
This paper introduces Channel Importance-driven learned Image Coding for Machines (CI-ICM), a novel image compression method tailored for machine vision tasks. CI-ICM uses a Channel Importance Generation (CIG) module to rank feature channels by importance, a Feature Channel Grouping and Scaling (FCGS) module to allocate bitrate non-uniformly, and a Task-Specific Channel Adaptation (TSCA) module to enhance features for downstream tasks. Experiments on COCO2017 demonstrate significant BD-mAP gains in object detection and instance segmentation compared to baseline codecs.
Machine vision tasks get up to 16% better performance from compressed images thanks to a new codec that prioritizes important feature channels.
Traditional human vision-centric image compression methods are suboptimal for machine vision centric compression due to different visual properties and feature characteristics. To address this problem, we propose a Channel Importance-driven learned Image Coding for Machines (CI-ICM), aiming to maximize the performance of machine vision tasks at a given bitrate constraint. First, we propose a Channel Importance Generation (CIG) module to quantify channel importance in machine vision and develop a channel order loss to rank channels in descending order. Second, to properly allocate bitrate among feature channels, we propose a Feature Channel Grouping and Scaling (FCGS) module that non-uniformly groups the feature channels based on their importance and adjusts the dynamic range of each group. Based on FCGS, we further propose a Channel Importance-based Context (CI-CTX) module to allocate bits among feature groups and to preserve higher fidelity in critical channels. Third, to adapt to multiple machine tasks, we propose a Task-Specific Channel Adaptation (TSCA) module to adaptively enhance features for multiple downstream machine tasks. Experimental results on the COCO2017 dataset show that the proposed CI-ICM achieves BD-mAP@50:95 gains of 16.25$\%$ in object detection and 13.72$\%$ in instance segmentation over the established baseline codec. Ablation studies validate the effectiveness of each contribution, and computation complexity analysis reveals the practicability of the CI-ICM. This work establishes feature channel optimization for machine vision-centric compression, bridging the gap between image coding and machine perception.
