Lightweight Multi-Vehicle Collaborative Perception Acceleration with Fusion Position Adjustment

2026-06-26

AI Summary

This work addresses the challenge of achieving both real-time performance and high accuracy in multi-vehicle cooperative perception under constrained computational and communication resources. To this end, the authors propose a lightweight intermediate fusion acceleration method termed Fusion position Adjustment between Linear Layers (FALL). By theoretically establishing the equivalence in perceptual accuracy between early and late fusion positions, and leveraging the conditional additivity property of linear layers in deep neural networks, FALL enables lossless optimization of feature fusion placement. Experimental results demonstrate that the proposed approach reduces latency by 74.8% under communication-constrained settings and by 30.3% under computation-constrained scenarios, substantially enhancing system efficiency without compromising accuracy.

Abstract

Multi-vehicle collaborative perception (MvCP) is considered as a key technology to facilitate automated driving (AD), where real-time MvCP under limited resources is significant for reliable AD. In this paper, we formulate a lightweight acceleration scheme for intermediate-fusion (IF) MvCP, which can adapt to both situations of limited computation and communication resources. We provide a relaxed definition conditional additivity and analyze the conditional additivity for various DNN linear layers. On this basis, we focus on the IF-MvCP based on additive feature fusion, and derive the MvCP precision consistency of the forward and backward feature fusion position (FP) adjustments among linear layers. Through experiments, we further validate the precision consistency of the FP adjustment method. Moreover, we propose an FP adjustment among linear layers (FALL) scheme for MvCP acceleration without precision loss theoretically. Simulation results show that the proposed FALL can reduce MvCP latency by up to 74.8% under limited communication resources and by up to 30.3% under limited computation resources.

Problem

• multi-vehicle collaborative perception
• lightweight acceleration
• resource constraints
• real-time perception
• precision consistency

Innovation

• multi-vehicle collaborative perception
• fusion position adjustment
• conditional additivity
• lightweight acceleration
• intermediate fusion