NBBOX: Noisy Bounding Box Improves Remote Sensing Object Detection

2024-09-14

arXiv.org

AI Summary

To address inconsistent annotation of rotated bounding boxes and limited detection performance under small-sample conditions in remote sensing imagery, this paper proposes NBBOX—a bounding-box-level noise augmentation method. Unlike conventional image-level augmentation, NBBOX applies controllable geometric perturbations (i.e., scaling, rotation, and translation) directly to ground-truth bounding boxes during training. It is the first systematic study of box-level noise injection, specifically designed for rotated object detection and modeling annotation uncertainty, without introducing auxiliary network modules or additional computational overhead. Extensive experiments on mainstream detectors—including Rotated Faster R-CNN and YOLOv8-R—demonstrate consistent improvements: mAP gains of 1.2–2.7% on DOTA and DIOR-R benchmarks, alongside ~15% faster training convergence. NBBOX significantly outperforms image-level augmentation strategies such as CutMix and Mosaic, establishing a more effective and efficient paradigm for robust rotated object detection in remote sensing.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Intelligent Robots: Learning & Optimization for ROB
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Representation, semantic annotation, enhancement, enrichments, access and/or integration of a variety of data on the Web

Abstract

Data augmentation has shown significant advancements in computer vision to improve model performance over the years, particularly in scenarios with limited and insufficient data. Currently, most studies focus on adjusting the image or its features to expand the size, quality, and variety of samples during training in various tasks including object detection. However, we argue that it is necessary to investigate bounding box transformations as a data augmentation technique rather than image-level transformations, especially in aerial imagery due to potentially inconsistent bounding box annotations. Hence, this letter presents a thorough investigation of bounding box transformation in terms of scaling, rotation, and translation for remote sensing object detection. We call this augmentation strategy NBBOX (Noise Injection into Bounding Box). We conduct extensive experiments on DOTA and DIOR-R, both well-known datasets that include a variety of rotated generic objects in aerial images. Experimental results show that our approach significantly improves remote sensing object detection without whistles and bells and it is more time-efficient than other state-of-the-art augmentation strategies.

Problem

• Object Recognition
• Aerial Imagery
• Scale Variation

Innovation

• NBBOX
• Data Augmentation
• Object Recognition