YOLO-PEFT：面向YOLO系列的结构感知参数高效微调框架
原标题：YOLO-PEFT: Parameter-Efficient Fine-Tuning on YOLO Family
AI 摘要
腾讯提出的YOLO-PEFT框架将适配器放置问题形式化为可审计的约束规划问题，为YOLO系列实时检测器提供参数高效微调方案。在VOC07+12基准上，RS-LoRA在YOLO11s和YOLO12s上分别达到0.7138和0.7307 mAP50-95，优于全量微调，同时LoRA可降低43.9%的峰值训练内存，但训练时间增加1.72倍。该框架支持在训练前返回拒绝决策，但未见架构的拒绝验证仍是开放问题。
正文节选
YOLO-PEFT: Parameter-Efficient Fine-Tuning on YOLO Family Abstract Generic parameter-efficient fine-tuning (PEFT) methods transferred from language models can fail silently on real-time detectors, whose heterogeneous operators and detection-specific components impose placement constraints absent from regular Transformer stacks. We propose YOLO-PEFT, a structure-aware framework that formulates adapter placement as an auditable constraint-planning problem. Given a detector graph, a PEFT request, a