Dynamic Belief Fusion for Object Detection

TL;DR

This paper introduces Dynamic Belief Fusion (DBF), a novel method that effectively combines multiple object detectors by estimating detection ambiguity and dynamically assigning probabilities, resulting in improved detection accuracy.

Contribution

The main novelty is the DBF method that dynamically fuses heterogeneous detector outputs using Dempster's rule based on confidence levels and prior performance.

Findings

DBF outperforms conventional fusion methods.

Detection accuracy is significantly improved.

Effective on ARL and PASCAL VOC 07 datasets.

Abstract

A novel approach for the fusion of heterogeneous object detection methods is proposed. In order to effectively integrate the outputs of multiple detectors, the level of ambiguity in each individual detection score is estimated using the precision/recall relationship of the corresponding detector. The main contribution of the proposed work is a novel fusion method, called Dynamic Belief Fusion (DBF), which dynamically assigns probabilities to hypotheses (target, non-target, intermediate state (target or non-target)) based on confidence levels in the detection results conditioned on the prior performance of individual detectors. In DBF, a joint basic probability assignment, optimally fusing information from all detectors, is determined by the Dempster's combination rule, and is easily reduced to a single fused detection score. Experiments on ARL and PASCAL VOC 07 datasets demonstrate that the detection accuracy of DBF is considerably greater than conventional fusion approaches as well as individual detectors used for the fusion.