Abstract:Aerial image objects are usually orientated arbitrarily, with a large scale range, and densely distributed. Traditional horizontal bounding box (HBB) detectors tend to filter out densely distributed objects leading to missed detections, such as ship (SH) and vehicle. Therefore, oriented object detection has become a mainstream solution in recent years. The 2-D Gaussian distribution representation of the oriented bounding boxes (OBBs) solves the problem of angular discontinuity and boundary discontinuity and thus gets more attention. However, as the aspect ratio of the object gradually decreases, its predicted angular performance continues to decrease. We find that the angular gradient of an object decreases sharply as the aspect ratio decreases, resulting in a large gradient gap between a small aspect ratio object (SARO) and a large aspect ratio object (LARO). It makes the detector prefer to ignore SARO during training, which weakens the high precision performance of SARO. We call this phenomenon shape imbalance. To solve the problem, we proposed a simple gradient rebalancing strategy named shape balance. Since the shape imbalance is only related to the aspect ratio of the object, we designed a modulation function with an inverse aspect ratio to calculate the balance coefficient. The principle of the function is that the larger the aspect ratio, the smaller the balance coefficient; the smaller the aspect ratio, the larger the balance coefficient. We aim to get the balance coefficients for objects with different aspect ratios. Location loss multiplied by a balance coefficient can directly adjust the gradient gap between objects with different aspect ratios to achieve a rebalancing effect. Extensive experiments conducted on DOTA-v1.0 dataset and DIOR-R dataset verify the effectiveness of our proposed method. Our method improves the detection performance of Gaussian location loss by an average of 2.08%/1.01%(AP75/mAP) metrics on the DOTA-v1.0 dataset and 1.17%/0.82%(AP75/mAP) improvements for DIOR-R dataset.

Overlapping object detection with adaptive Gaussian sample division and asymmetric weighted loss

Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection

Local–Global Attentive Adaptation for Object Detection

Cross-domain Object Detection by Local to Global Object-Aware Feature Alignment

Gaussian similarity-based adaptive dynamic label assignment for tiny object detection

LapNet : Automatic Balanced Loss and Optimal Assignment for Real-Time Dense Object Detection

Revisiting the Loss Weight Adjustment in Object Detection

Dynamic Anchor Learning for Arbitrary-Oriented Object Detection

Balance-Oriented Focal Loss with Linear Scheduling for Anchor Free Object Detection

Balanced One-Stage Object Detection by Enhancing the Effect of Positive Samples

Class-Aware Prediction: A Solution to Center Point Collision for Anchor-Free Object Detection in Aerial Images

Revisiting Foreground-Background Imbalance in Object Detectors

Oriented Object Detector with Gaussian Distribution Cost Label Assignment and Task-Decoupled Head

Rebalancing Gaussian Location Loss for High-Precision Detection on Remote Sensing Images

Dense Object Detection Based on De-Homogenized Queries

Gaussian Aware Anchor-Free Rotated Detector for Aerial Object Detection

Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection

PICA: Point-wise Instance and Centroid Alignment Based Few-shot Domain Adaptive Object Detection with Loose Annotations

Adaptive Scale and Spatial Aggregation for Real-Time Object Detection

Hybrid Classification-Regression Adaptive Loss for Dense Object Detection

Partial Alignment for Object Detection in the Wild