无需训练，用背景嵌入记忆抑制实时监控中的误检。
BEM: Training-Free Background Embedding Memory for False-Positive Suppression in Real-Time Fixed-Background Camera
- 利用稳定背景构建无标签嵌入记忆，推理时动态修正检测分数。
- 在LLVIP和模拟监控数据上，误检率显著下降，召回率几乎不变。
- 适合部署于固定摄像头场景的实时目标检测系统，无需重新训练。
预训练检测器在基准测试中表现良好，但在实际部署中常因训练数据与真实环境分布差异导致性能下降。现有COCO类基准侧重类别多样性而非实例密度，使基于每类稀疏训练的检测器在监控、交通等密集单类或少类场景中表现不佳。在固定摄像头环境中，准静态背景可作为无标签先验，在推理时用于抑制虚假检测。为此，我们提出背景嵌入记忆（BEM），一种轻量级、无需训练、权重冻结的模块，可直接接入预训练检测器进行推理。BEM估计干净背景嵌入，维护原型记忆，并通过逆相似性、秩加权惩罚重新评分检测置信度，有效降低误检率同时保持召回率。实验发现，背景帧余弦相似度与物体数量呈负相关，与精度-置信度AUC（P-AUC）正相关，支持其作为免训练控制信号。在YOLO与RT-DETR系列模型上，于LLVIP及模拟监控流中，BEM持续减少误检，且保持实时性能。代码已开源：https://github.com/Leo-Park1214/Background-Embedding-Memory.git。
原文摘要 · Abstract (English)
Pretrained detectors perform well on benchmarks but often suffer performance degradation in real-world deployments due to distribution gaps between training data and target environments. COCO-like benchmarks emphasize category diversity rather than instance density, causing detectors trained under per-class sparsity to struggle in dense, single- or few-class scenes such as surveillance and traffic monitoring. In fixed-camera environments, the quasi-static background provides a stable, label-free prior that can be exploited at inference to suppress spurious detections. To address the issue, we propose Background Embedding Memory (BEM), a lightweight, training-free, weight-frozen module that can be attached to pretrained detectors during inference. BEM estimates clean background embeddings, maintains a prototype memory, and re-scores detection logits with an inverse-similarity, rank-weighted penalty, effectively reducing false positives while maintaining recall. Empirically, background-frame cosine similarity correlates negatively with object count and positively with Precision-Confidence AUC (P-AUC), motivating its use as a training-free control signal. Across YOLO and RT-DETR families on LLVIP and simulated surveillance streams, BEM consistently reduces false positives while preserving real-time performance. Our code is available at https://github.com/Leo-Park1214/Background-Embedding-Memory.git
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。