Embodied Image Compression

TL;DR

This paper introduces the novel concept of Embodied Image Compression, addressing the unique challenges of compressing visual data for Embodied AI agents in real-world environments, and provides a benchmark for evaluation.

Contribution

It establishes the first scientific framework and benchmark, EmbodiedComp, for evaluating image compression specifically designed for Embodied AI in real-world scenarios.

Findings

Existing VLAs fail at low bitrates in embodied tasks

EmbodiedComp enables systematic evaluation of compression methods

Empirical results highlight the need for domain-specific compression techniques

Abstract

Image Compression for Machines (ICM) has emerged as a pivotal research direction in the field of visual data compression. However, with the rapid evolution of machine intelligence, the target of compression has shifted from task-specific virtual models to Embodied agents operating in real-world environments. To address the communication constraints of Embodied AI in multi-agent systems and ensure real-time task execution, this paper introduces, for the first time, the scientific problem of Embodied Image Compression. We establish a standardized benchmark, EmbodiedComp, to facilitate systematic evaluation under ultra-low bitrate conditions in a closed-loop setting. Through extensive empirical studies in both simulated and real-world settings, we demonstrate that existing Vision-Language-Action models (VLAs) fail to reliably perform even simple manipulation tasks when compressed below the Embodied bitrate threshold. We anticipate that EmbodiedComp will catalyze the development of domain-specific compression tailored for Embodied agents , thereby accelerating the Embodied AI deployment in the Real-world.