用低秩微调提升小模型在航拍图像少样本检测中的适应能力
Analyzing the Impact of Low-Rank Adaptation for Cross-Domain Few-Shot Object Detection in Aerial Images
- 将LoRA引入DiffusionDet，实现资源受限下的高效微调
- 1-shot和5-shot下性能略有提升，高样本量时全微调更优
- 适合关注少样本学习与参数效率的科研人员
本文研究了低秩适应（LoRA）在航拍图像跨域少样本目标检测中小模型中的应用。原为大规模模型设计的LoRA有助于缓解过拟合，适用于资源受限场景。我们将LoRA集成至DiffusionDet，分别在DOTA和DIOR数据集上评估其性能。结果表明，在初始微调后应用LoRA可轻微提升低样本设置（如1-shot和5-shot）下的表现，而在更高样本量配置中，全微调仍更具优势。这些发现凸显了LoRA在航拍目标检测中高效适配的潜力，推动了参数高效微调策略在少样本学习中的进一步研究。代码已公开：https://github.com/HichTala/LoRA-DiffusionDet。
原文摘要 · Abstract (English)
This paper investigates the application of Low-Rank Adaptation (LoRA) to small models for cross-domain few-shot object detection in aerial images. Originally designed for large-scale models, LoRA helps mitigate overfitting, making it a promising approach for resource-constrained settings. We integrate LoRA into DiffusionDet, and evaluate its performance on the DOTA and DIOR datasets. Our results show that LoRA applied after an initial fine-tuning slightly improves performance in low-shot settings (e.g., 1-shot and 5-shot), while full fine-tuning remains more effective in higher-shot configurations. These findings highlight LoRA's potential for efficient adaptation in aerial object detection, encouraging further research into parameter-efficient fine-tuning strategies for few-shot learning. Our code is available here: https://github.com/HichTala/LoRA-DiffusionDet.
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。