YOLO-DS通过双统计协同解耦特征，提升目标检测精度。
YOLO-DS: Fine-Grained Feature Decoupling via Dual-Statistic Synergy Operator for Object Detection
- 用通道均值与峰均差联合建模，解耦异质物体特征
- 在COCO上5种规模模型均提升1.1%~1.7%的AP，延迟几乎不变
- 适合追求高精度且对推理速度敏感的实时检测场景
单阶段目标检测（尤其是YOLO系列）在精度与效率间取得了良好平衡。然而，现有YOLO检测器缺乏对共享特征通道内异质物体响应的显式建模，限制了性能进一步提升。为此，我们提出YOLO-DS，其核心为新型双统计协同算子（DSO）。DSO通过联合建模通道均值与峰均差，实现物体特征的解耦。基于DSO，设计两个轻量级门控模块：用于自适应通道选择的双统计协同门控（DSG）模块，以及用于深度维度特征加权的多路径分段门控（MSG）模块。在MS-COCO基准上，YOLO-DS在五种模型尺度（N, S, M, L, X）下均优于YOLOv8，AP提升1.1%至1.7%，且推理延迟增加极小。大量可视化、消融与对比实验验证了方法的有效性，表明其在高效区分异质物体方面具有显著优势。
原文摘要 · Abstract (English)
One-stage object detection, particularly the YOLO series, strikes a favorable balance between accuracy and efficiency. However, existing YOLO detectors lack explicit modeling of heterogeneous object responses within shared feature channels, which limits further performance gains. To address this, we propose YOLO-DS, a framework built around a novel Dual-Statistic Synergy Operator (DSO). The DSO decouples object features by jointly modeling the channel-wise mean and the peak-to-mean difference. Building upon the DSO, we design two lightweight gating modules: the Dual-Statistic Synergy Gating (DSG) module for adaptive channel-wise feature selection, and the Multi-Path Segmented Gating (MSG) module for depth-wise feature weighting. On the MS-COCO benchmark, YOLO-DS consistently outperforms YOLOv8 across five model scales (N, S, M, L, X), achieving AP gains of 1.1% to 1.7% with only a minimal increase in inference latency. Extensive visualization, ablation, and comparative studies validate the effectiveness of our approach, demonstrating its superior capability in discriminating heterogeneous objects with high efficiency.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。