Long-SCOPE: Fully Sparse Long-Range Cooperative 3D Perception

arXiv cs.CV / 4/13/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

Long-SCOPE is presented as a fully sparse framework for cooperative 3D perception using V2X communication, aiming to extend sensing range and resolve occlusions in autonomous driving.
The work targets two main deployment bottlenecks in existing approaches: quadratic scaling from dense BEV representations and brittle feature association under large observation/alignment errors.
It introduces a Geometry-guided Query Generation module to improve detection of small, distant objects and a learnable Context-Aware Association module to match queries robustly despite severe positional noise.
Experiments on the V2X-Seq and Griffin datasets report state-of-the-art results, with particular gains in difficult 100–150 m long-range scenarios, while keeping computation and communication costs competitive.

Abstract

Cooperative 3D perception via Vehicle-to-Everything communication is a promising paradigm for enhancing autonomous driving, offering extended sensing horizons and occlusion resolution. However, the practical deployment of existing methods is hindered at long distances by two critical bottlenecks: the quadratic computational scaling of dense BEV representations and the fragility of feature association mechanisms under significant observation and alignment errors. To overcome these limitations, we introduce Long-SCOPE, a fully sparse framework designed for robust long-distance cooperative 3D perception. Our method features two novel components: a Geometry-guided Query Generation module to accurately detect small, distant objects, and a learnable Context-Aware Association module that robustly matches cooperative queries despite severe positional noise. Experiments on the V2X-Seq and Griffin datasets validate that Long-SCOPE achieves state-of-the-art performance, particularly in challenging 100-150 m long-range settings, while maintaining highly competitive computation and communication costs.