🤖 AI Summary
This work addresses the challenge of cross-rig generalization in autonomous driving perception systems, which are typically trained under fixed sensor configurations and struggle to adapt to diverse real-world camera layouts due to geometric domain shifts. To this end, the authors introduce the Plentiful CARLA Camera Rigs benchmark, which renders data from 14 systematically designed camera rigs within identical driving scenes. They propose the first controllable evaluation framework for cross-rig generalization and introduce two calibration-based geometric discrepancy metrics—Rig Variance and Rig Contrastive Distance—to quantify inter-rig differences and transfer difficulty. Experimental results demonstrate a strong correlation between geometric discrepancy and performance degradation, with Rig Contrastive Distance effectively predicting the relative difficulty of cross-rig transfer.
📝 Abstract
Camera-based perception systems for autonomous driving are typically developed and evaluated using fixed sensor rigs, while real-world vehicle fleets exhibit substantial variation in camera placement, orientation, field of view, and camera count. This mismatch introduces a cross-rig domain gap in which only the geometric observation process changes. To study this effect under controlled conditions, we introduce Plentiful CARLA Camera Rigs, a benchmark that renders identical driving scenes under 14 systematically designed camera rigs. This setup enables direct analysis of cross-rig generalization without confounding changes in scene content or appearance. Using the benchmark, we analyze cross-rig transfer behavior of representative multi-view perception architectures and observe substantial performance shifts induced by geometric rig variation. To facilitate structured analysis, we further introduce two calibration-based descriptors derived from rig metadata: Rig Variance, capturing internal rig diversity, and Rig Contrastive Distance, measuring geometric discrepancy between rigs. Our experiments show that geometric rig differences strongly correlate with relative cross-rig performance shifts and that Rig Contrastive Distance provides a reliable proxy for ranking transfer difficulty between sensor rigs.