Learnability-Driven Submodular Optimization for Active Roadside 3D Detection

TL;DR

This paper introduces LH3D, an active learning framework for roadside monocular 3D detection that prioritizes learnability over uncertainty, reducing annotation effort while maintaining high detection performance.

Contribution

It proposes a learnability-driven scene selection method for active learning in roadside 3D detection, effectively suppressing ambiguous samples and improving annotation efficiency.

Findings

Achieves over 86% of full performance with only 25% annotation budget

Outperforms uncertainty-based active learning baselines

Demonstrates learnability is more critical than uncertainty in this context

Abstract

Roadside perception datasets are typically constructed via cooperative labeling between synchronized vehicle and roadside frame pairs. However, real deployment often requires annotation of roadside-only data due to hardware and privacy constraints. Even human experts struggle to produce accurate labels without vehicle-side data (image, LIDAR), which not only increases annotation difficulty and cost, but also reveals a fundamental learnability problem: many roadside-only scenes contain distant, blurred, or occluded objects whose 3D properties are ambiguous from a single view and can only be reliably annotated by cross-checking paired vehicle--roadside frames. We refer to such cases as inherently ambiguous samples. To reduce wasted annotation effort on inherently ambiguous samples while still obtaining high-performing models, we turn to active learning. This work focuses on active learning for roadside monocular 3D object detection and proposes a learnability-driven framework that selects scenes which are both informative and reliably labelable, suppressing inherently ambiguous samples while ensuring coverage. Experiments demonstrate that our method, LH3D, achieves 86.06%, 67.32%, and 78.67% of full-performance for vehicles, pedestrians, and cyclists respectively, using only 25% of the annotation budget on DAIR-V2X-I, significantly outperforming uncertainty-based baselines. This confirms that learnability, not uncertainty, matters for roadside 3D perception.