Semantic-decoupled Spatial Partition Guided Point-supervised Oriented Object Detection

2025-06-12

AI Summary

To address coarse sample assignment and instance ambiguity in point-supervised oriented object detection for high-density remote sensing scenes, this paper proposes the Semantic-Decoupled Spatial Partitioning (SSP) framework. Methodologically, SSP introduces: (1) a novel pixel-level spatial partitioning mechanism for fine-grained positive/negative sample mining; (2) semantic map modulation and pixel-wise evaluation, integrating prior-guided and data-driven label purification to enhance pseudo-label reliability; and (3) a semantic-modulated spatial partitioning box extraction strategy to mitigate instance confusion caused by rigid partitioning rules. Evaluated under the DOTA-v1.0 point-supervision setting, SSP achieves 45.78% mAP, outperforming the state-of-the-art by 4.10%. When integrated with ORCNN and ReDet, performance further improves to 47.86% and 48.50% mAP, respectively.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Semi-Supervised Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services

Abstract

Recent remote sensing tech advancements drive imagery growth, making oriented object detection rapid development, yet hindered by labor-intensive annotation for high-density scenes. Oriented object detection with point supervision offers a cost-effective solution for densely packed scenes in remote sensing, yet existing methods suffer from inadequate sample assignment and instance confusion due to rigid rule-based designs. To address this, we propose SSP (Semantic-decoupled Spatial Partition), a unified framework that synergizes rule-driven prior injection and data-driven label purification. Specifically, SSP introduces two core innovations: 1) Pixel-level Spatial Partition-based Sample Assignment, which compactly estimates the upper and lower bounds of object scales and mines high-quality positive samples and hard negative samples through spatial partitioning of pixel maps. 2) Semantic Spatial Partition-based Box Extraction, which derives instances from spatial partitions modulated by semantic maps and reliably converts them into bounding boxes to form pseudo-labels for supervising the learning of downstream detectors. Experiments on DOTA-v1.0 and others demonstrate SSP' s superiority: it achieves 45.78% mAP under point supervision, outperforming SOTA method PointOBB-v2 by 4.10%. Furthermore, when integrated with ORCNN and ReDet architectures, the SSP framework achieves mAP values of 47.86% and 48.50%, respectively. The code is available at https://github.com/antxinyuan/ssp.

Problem

• Reducing labor-intensive annotation in high-density oriented object detection
• Improving sample assignment and instance confusion in point-supervised methods
• Enhancing pseudo-label quality for downstream detector learning

Innovation

• Pixel-level spatial partition for sample assignment
• Semantic spatial partition for box extraction
• Rule-driven and data-driven unified framework