CertainNet: Sampling-free Uncertainty Estimation for Object Detection

TL;DR

CertainNet introduces a novel, sampling-free method for estimating uncertainties in object detection, providing separate, calibrated uncertainty measures for each output component, enhancing safety in autonomous driving.

Contribution

It is the first to offer separate uncertainty estimates for objectness, class, location, and size in object detection, using an uncertainty-aware heatmap and neighboring bounding boxes.

Findings

Outperforms previous methods and ensembles in out-of-domain scenarios.

Provides real-time, comprehensive uncertainty estimates.

Generalizes better to unseen datasets.

Abstract

Estimating the uncertainty of a neural network plays a fundamental role in safety-critical settings. In perception for autonomous driving, measuring the uncertainty means providing additional calibrated information to downstream tasks, such as path planning, that can use it towards safe navigation. In this work, we propose a novel sampling-free uncertainty estimation method for object detection. We call it CertainNet, and it is the first to provide separate uncertainties for each output signal: objectness, class, location and size. To achieve this, we propose an uncertainty-aware heatmap, and exploit the neighboring bounding boxes provided by the detector at inference time. We evaluate the detection performance and the quality of the different uncertainty estimates separately, also with challenging out-of-domain samples: BDD100K and nuImages with models trained on KITTI. Additionally, we propose a new metric to evaluate location and size uncertainties. When transferring to unseen datasets, CertainNet generalizes substantially better than previous methods and an ensemble, while being real-time and providing high quality and comprehensive uncertainty estimates.