Uncertainty Quantification of Collaborative Detection for Self-Driving

Abstract

Sharing information between connected and autonomous vehicles (CAVs) fundamentally improves the performance of collaborative object detection for self-driving. However, CAVs still have uncertainties on object detection due to practical challenges, which will affect the later modules in self-driving such as planning and control. Hence, uncertainty quantification is crucial for safety-critical systems such as CAVs. Our work is the first to estimate the uncertainty of collaborative object detection. We propose a novel uncertainty quantification method, called Double-M Quantification, which tailors a moving block bootstrap (MBB) algorithm with direct modeling of the multivariant Gaussian distribution of each corner of the bounding box. Our method captures both the epistemic uncertainty and aleatoric uncertainty with one inference pass based on the offline Double-M training process. And it can be used with different collaborative object detectors. Through experiments on the comprehensive collaborative perception dataset, we show that our Double-M method achieves more than 4× improvement on uncertainty score and more than 3% accuracy improvement, compared with the state-of-the-art uncertainty quantification methods. Our code is public on https://coperception. github.io/double-m-quantification/.

Collaborative Perception in Autonomous Driving: Methods, Datasets and Challenges

“…Besides, the collaboration among agents will bring redundant and uncertain information. Accordingly, it is necessary to capture this latent information [29,54]. Because existing collaborative perception methods are built with different collaboration modules and diverse strategies, we summarize these modules according to their stage, as shown in Fig.…”

“…Besides redundancy, the connected autonomous vehicle also contains perceptual uncertainties, which reflect perception inaccuracy or sensor noises. Su et al [54] firstly explore the uncertainty in collaboration perception. Specifically, they design Double-M Quantification tailors moving block bootstrap to estimate both the model and data uncertainty, together with a well-designed direct modeling component.…”

Cooperative control method of multi-missile formation under uncontrollable speed

“…have as a group by exchanging information with each other so as to use their combined sensory experiences to perceive. Along this direction, recent efforts have been made to deliver datasets [39,38,14,5] and cooperative solutions [21,20,1,5,27,30,10,2,33,37,38].…”