UniTAC: Universal Task-Aware Compression via Weighted Distortion Measures

2026-08-17

AI Summary

This study addresses the vulnerability of specialized encoders in bandwidth-constrained physical AI systems caused by dynamic downstream tasks. We propose UniTAC, a universal task-aware codec that conditions a Vision Transformer via runtime injection of task importance vectors and establishes a weighted rate-distortion framework. This approach enables seamless switching between universal and specialized modes without retraining, facilitating efficient multi-task compression. Experimental results demonstrate that at 0.034 bpp, UniTAC achieves 91.4% accuracy, trailing specialized encoders by only 1.9% while significantly outperforming universal baselines by 14.5 percentage points. These findings validate UniTAC as an effective solution for adapting a single model to diverse tasks under strict bandwidth limitations, successfully bridging the performance gap between general-purpose and task-specific compression paradigms.

Abstract

Physical AI systems such as autonomous vehicles and robots rely on timely exchange of high-dimensional sensory signals under tight bandwidth, latency, and energy budgets. Because the task driving downstream decisions evolves over time, a task-specific codec is brittle and retraining one per task is infeasible in the field. We propose UniTAC, a single learned image codec spanning universal (task-agnostic) to task-specialized operation, re-targeted at runtime without retraining. The task is abstracted as a per-component importance vector, derived, e.g., from gradient attribution of any downstream model, and transmitted as low-overhead side information that conditions both encoder and decoder. Trained once over a broad, randomized family of such vectors against weighted-reconstruction distortion, UniTAC keeps a fixed backbone and a single human-viewable reconstruction whose fidelity is steered to the active task by swapping the injected vector. We analyze the underlying weighted rate-distortion problem, characterizing when a diagonal weighted distortion is task-consistent and how weights relate to task sensitivity. Guided by this, we design a Vision Transformer (ViT) codec whose token-level conditioning natively realizes this weight-driven code. On a localized task at 0.034 bpp, a single UniTAC model reaches 91.4% accuracy, only 1.9% below a task-based codec (93.3%) and above universal codecs (76.9%).

Problem

• Task-Aware Compression
• Physical AI Systems
• Dynamic Task Adaptation
• Learned Image Codec
• Weighted Distortion

Innovation

• Universal Task-Aware Compression
• Weighted Distortion Measures
• Runtime Retargeting
• Token-Level Conditioning
• Vision Transformer Codec