Mutual Information-calibrated Conformal Feature Fusion for Uncertainty-Aware Multimodal 3D Object Detection at the Edge

TL;DR

This paper introduces a novel multimodal 3D object detection framework that combines conformal inference and mutual information to provide lightweight, uncertainty-aware predictions suitable for real-time edge robotics.

Contribution

It integrates conformal inference with mutual information to calibrate uncertainty bounds in a multimodal VAE-based detection framework, enhancing robustness and real-time applicability.

Findings

Inverse correlation between uncertainty and NMI during training

Comparable or superior detection performance on KITTI benchmarks

Effective uncertainty calibration without Monte Carlo sampling

Abstract

In the expanding landscape of AI-enabled robotics, robust quantification of predictive uncertainties is of great importance. Three-dimensional (3D) object detection, a critical robotics operation, has seen significant advancements; however, the majority of current works focus only on accuracy and ignore uncertainty quantification. Addressing this gap, our novel study integrates the principles of conformal inference (CI) with information theoretic measures to perform lightweight, Monte Carlo-free uncertainty estimation within a multimodal framework. Through a multivariate Gaussian product of the latent variables in a Variational Autoencoder (VAE), features from RGB camera and LiDAR sensor data are fused to improve the prediction accuracy. Normalized mutual information (NMI) is leveraged as a modulator for calibrating uncertainty bounds derived from CI based on a weighted loss function. Our simulation results show an inverse correlation between inherent predictive uncertainty and NMI throughout the model's training. The framework demonstrates comparable or better performance in KITTI 3D object detection benchmarks to similar methods that are not uncertainty-aware, making it suitable for real-time edge robotics.