arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2407.07586cs.CVcs.LG
收藏

简化目标检测的无源域自适应:有效的自训练策略与性能洞察

Simplifying Source-Free Domain Adaptation for Object Detection: Effective Self-Training Strategies and Performance Insights

EPFL(洛桑联邦理工学院)

Yan Hao, Florent Forest, Olga Fink

AI总结:

针对无源目标检测任务,研究提出仅适配批量统计量的强基线、无源无偏教师方法及固定伪标签训练策略,在多个驾驶数据集自适应任务上取得优于现有复杂方法的性能,且更高效、可缓解师生崩塌问题。

摘要:

本文聚焦计算机视觉中目标检测的无源域自适应任务。由于为每个新域获取标注数据集的成本高昂,该任务极具挑战性且具备很高的实用价值。近期研究针对无源目标检测(Source-Free Object Detection,SFOD)提出了多种解决方案,大多是教师-学生架构的变体,搭配各类特征对齐、正则化和伪标签选择策略。\n本研究探究了更简单的方法,并在多种自适应场景下对比了其与更复杂SFOD方法的性能。我们强调了检测器骨干网络中批量归一化层的重要性,并证明仅适配批量统计量就是SFOD的强基线。我们提出了无源场景下结合强弱增强的均值教师的简单扩展——无源无偏教师(Source-Free Unbiased Teacher,SF-UT),并证明其性能实际上优于大多数现有SFOD方法。此外,我们还展示了一种更简单的策略:基于固定伪标签集进行训练,该策略能达到与更复杂的师生互学习相近的性能,同时计算效率更高,还能缓解师生崩塌这一主要问题。\n我们使用(Foggy)Cityscapes、Sim10k和KITTI等驾驶基准数据集,在多个自适应任务上开展了实验,在Cityscapes→Foggy-Cityscapes任务上,相比最新的SFOD最优方法,AP50实现了4.7%的显著提升。源代码可在https://github.com/EPFL-IMOS/simple-SFOD获取。

英文摘要:

This paper focuses on source-free domain adaptation for object detection in computer vision. This task is challenging and of great practical interest, due to the cost of obtaining annotated data sets for every new domain. Recent research has proposed various solutions for Source-Free Object Detection (SFOD), most being variations of teacher-student architectures with diverse feature alignment, regularization and pseudo-label selection strategies. Our work investigates simpler approaches and their performance compared to more complex SFOD methods in several adaptation scenarios. We highlight the importance of batch normalization layers in the detector backbone, and show that adapting only the batch statistics is a strong baseline for SFOD. We propose a simple extension of a Mean Teacher with strong-weak augmentation in the source-free setting, Source-Free Unbiased Teacher (SF-UT), and show that it actually outperforms most of the previous SFOD methods. Additionally, we showcase that an even simpler strategy consisting in training on a fixed set of pseudo-labels can achieve similar performance to the more complex teacher-student mutual learning, while being computationally efficient and mitigating the major issue of teacher-student collapse. We conduct experiments on several adaptation tasks using benchmark driving datasets including (Foggy)Cityscapes, Sim10k and KITTI, and achieve a notable improvement of 4.7\% AP50 on Cityscapes$\rightarrow$Foggy-Cityscapes compared with the latest state-of-the-art in SFOD. Source code is available at https://github.com/EPFL-IMOS/simple-SFOD.

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。

↑