Multiclass Confidence and Localization Calibration for Object Detection
This addresses calibration issues in object detection for security-sensitive and safety-critical applications, representing a novel extension beyond classification tasks.
The paper tackles the problem of overconfident predictions in object detection by proposing a train-time technique that jointly calibrates multiclass confidence and box localization using predictive uncertainties, achieving consistent reductions in calibration error across in-domain and out-of-domain benchmarks.
Albeit achieving high predictive accuracy across many challenging computer vision problems, recent studies suggest that deep neural networks (DNNs) tend to make overconfident predictions, rendering them poorly calibrated. Most of the existing attempts for improving DNN calibration are limited to classification tasks and restricted to calibrating in-domain predictions. Surprisingly, very little to no attempts have been made in studying the calibration of object detection methods, which occupy a pivotal space in vision-based security-sensitive, and safety-critical applications. In this paper, we propose a new train-time technique for calibrating modern object detection methods. It is capable of jointly calibrating multiclass confidence and box localization by leveraging their predictive uncertainties. We perform extensive experiments on several in-domain and out-of-domain detection benchmarks. Results demonstrate that our proposed train-time calibration method consistently outperforms several baselines in reducing calibration error for both in-domain and out-of-domain predictions. Our code and models are available at https://github.com/bimsarapathiraja/MCCL.