Abstract:Oriented object detection is a challenging and relatively new problem. Most existing approaches are based on deep learning and explore Oriented Bounding Boxes (OBBs) to represent the objects. They are typically based on adaptations of traditional detectors that work with Horizontal Bounding Boxes (HBBs), which have been exploring IoU-like loss functions to regress the HBBs. However, extending this idea for OBBs is challenging due to complex formulations or requirement for customized backpropagation implementations. Furthermore, using OBBs presents limitations for irregular or roughly circular objects, since the definition of the ideal OBB is an ambiguous and ill-posed problem. In this work, we jointly tackle the problem of training, representing, and evaluating oriented detectors. We explore Gaussian distributions–called Gaussian Bounding Boxes (GBBs)–as fuzzy representations for oriented objects and propose using a similarity metric between two GBBs based on the Hellinger distance. We show that this metric leads to a differentiable closed-form expression that can be directly used as a localization loss term to train OBB object detectors. We also show that GBBs present a natural representation as elliptical regions (called EBBs), which inherently mitigate ambiguity representation for circular objects. Finally, we empirically show that the proposed similarity metric computed between two GBBs strongly correlates with the IoU between the corresponding EBBs, motivating the name Probabilistic Intersection-over-Union (ProbIoU). Our experiments show that results using ProbIoU as a regression loss are competitive with state-of-the-art alternatives without requiring additional hyperparameters or customized implementations, and that ProbIoU is a promising alternative to evaluate oriented object detectors. Our code is available at https://github.com/ProbIOU/.

Unbiased IoU for Spherical Image Object Detection

Spherical Criteria for Fast and Accurate 360° Object Detection

Calibration of Central Omnidirectional Cameras Via the Viewing Sphere

Leveraging Front and Side Cues for Occlusion Handling in Monocular 3D Object Detection

3D IoU-Net: IoU Guided 3D Object Detector for Point Clouds

Rethinking IoU-based Optimization for Single-stage 3D Object Detection

Optimization for Anchor-Free Object Detection via Scale-Independent GIoU Loss

S-VoteNet: Deep Hough Voting with Spherical Proposal for 3D Object Detection.

Gaussian guided IoU: A better metric for balanced learning on object detection

Rethinking the Non-Maximum Suppression Step in 3D Object Detection from a Bird's-Eye View

Unified-IoU: For High-Quality Object Detection

3DIoUMatch: Leveraging IoU Prediction for Semi-Supervised 3D Object Detection

Constrained-SIoU: A Metric for Horizontal Candidates in Multi-Oriented Object Detection

IoU-aware Single-stage Object Detector for Accurate Localization

N-IoU: better IoU-based bounding box regression loss for object detection

Autonomous intersection over union (IoU) loss: adaptive dynamic non-monotonic focal IoU loss

OCM3D: Object-Centric Monocular 3D Object Detection

Probabilistic Intersection-Over-Union for Training and Evaluation of Oriented Object Detectors

EC-IoU: Orienting Safety for Object Detectors via Ego-Centric Intersection-over-Union

Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression