强而简洁：一种基于CLIP的迁移学习实现域泛化密集感知的基线方法
Strong but simple: A Baseline for Domain Generalized Dense Perception by CLIP-based Transfer Learning
AI总结：
针对深度神经网络感知任务的域泛化难题，该研究提出仅对视觉语言预训练模型进行简单微调的方法，在多个密集感知基准上取得SOTA效果，挑战了传统ImageNet迁移学习的标准地位。
摘要：
域泛化（DG）仍是深度神经网络（DNNs）感知任务面临的重大挑战，合成数据、光照、天气或位置变化都会引发域偏移。视觉语言模型（VLMs）大幅提升了泛化能力，已被应用于各类任务。近期已有首批研究将VLMs用于域泛化分割与目标检测，并取得了出色的泛化效果。但这类方法均依赖复杂模块、特征增强框架或额外模型。与之形成鲜明对比的是，我们发现对视觉语言预训练模型进行简单微调，就能取得具有竞争力甚至更优的泛化结果，且应用方式极为简便。此外我们还发现，视觉语言预训练的泛化效果始终优于此前主流的纯视觉预训练，这对基于ImageNet的迁移学习作为域泛化标准方案的地位提出了挑战。在合成数据集GTA5上训练时，全量微调视觉语言预训练模型可达到域泛化SOTA水平。我们还在一个新的合成到真实基准的目标检测任务中验证了这一结论。在常用的Cityscapes到ACDC基准测试中，该方法取得了77.9%的mIoU，展现出更优的泛化能力。我们还发现其域内泛化效果也有所提升，在Cityscapes测试集上达到86.4% mIoU的新SOTA，位列排行榜第一。
英文摘要：
Domain generalization (DG) remains a significant challenge for perception based on deep neural networks (DNNs), where domain shifts occur due to synthetic data, lighting, weather, or location changes. Vision-language models (VLMs) marked a large step for the generalization capabilities and have been already applied to various tasks. Very recently, first approaches utilized VLMs for domain generalized segmentation and object detection and obtained strong generalization. However, all these approaches rely on complex modules, feature augmentation frameworks or additional models. Surprisingly and in contrast to that, we found that simple fine-tuning of vision-language pre-trained models yields competitive or even stronger generalization results while being extremely simple to apply. Moreover, we found that vision-language pre-training consistently provides better generalization than the previous standard of vision-only pre-training. This challenges the standard of using ImageNet-based transfer learning for domain generalization. Fully fine-tuning a vision-language pre-trained model is capable of reaching the domain generalization SOTA when training on the synthetic GTA5 dataset. Moreover, we confirm this observation for object detection on a novel synthetic-to-real benchmark. We further obtain superior generalization capabilities by reaching 77.9% mIoU on the popular Cityscapes-to-ACDC benchmark. We also found improved in-domain generalization, leading to an improved SOTA of 86.4% mIoU on the Cityscapes test set marking the first place on the leaderboard.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
