面向域自适应目标检测的专家—教师—学生协同学习
Expert-Teacher-Student Collaborative Learning for Domain Adaptive Object Detection
把大规模视觉基础模型当作免训练的「专家」，与教师模型协同生成伪标签并做类级原型对齐，为跨域检测提供了一条无需额外训练成本就能引入 VFM 泛化知识的路径，可直接迁移到其他需要伪标签监督的域自适应视觉任务。
原文依据 leverages the complementary knowledge of expert and teacher models to collaboratively generate high-quality pseudo labels
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论域自适应目标检测（DAOD）旨在将在源域上训练的目标检测器泛化到目标域，而域间差异会削弱其自适应性。近年来，在大规模网络数据集上预训练的视觉基础模型（VFM）展现出强大的泛化能力，许多方法借助它们来弥合域间差异。然而，其泛化知识并未针对特定域进行定制，因而难以在目标域中提供精确的指导。本文提出一种专家—教师—学生协同学习（ETS）框架，将来自 VFM 的泛化知识与来自教师模型的域特定知识协同起来。具体而言，我们首先设计专家—教师协同教学（ETCT）模块，利用专家模型与教师模型的互补知识协同生成高质量伪标签，以监督学生模型的学习。其次，我们设计专家—教师联合巩固（ETJC）模块，在专家、教师与学生模型之间引入类级原型对齐，以在学生模型中联合巩固泛化知识与域特定知识。ETS 以「免费午餐」的方式将 VFM 用作专家模型，从而避免了显著的额外训练成本。大量实验表明，我们的方法在三个基准上优于现有最先进方法。我们的代码见补充材料。
英文原文摘要 1500 字符
Domain adaptive object detection (DAOD) aims to generalize an object detector trained on a source domain to a target domain, where the domain gap degrades the adaptability. Recently, large-scale vision foundation models (VFMs), pretrained on web-scale datasets, exhibit such powerful generalization capabilities that many approaches leverage them to bridge the domain gap. However, their generalized knowledge is not tailored to the specific domain, which makes it difficult to offer precise guidance in the target domain. In this paper, we propose an Expert-Teacher-Student collaborative learning (ETS) framework to synergize the generalized knowledge from VFMs with the domain-specific knowledge from the teacher model. Concretely, we first design an Expert-Teacher Collaborative Teaching (ETCT) module, which leverages the complementary knowledge of expert and teacher models to collaboratively generate high-quality pseudo labels for supervising student model learning. Second, we devise an Expert-Teacher Joint Consolidating (ETJC) module, which introduces class-wise prototype alignment among expert, teacher, and student models, to jointly consolidate generalized and domain-specific knowledge within the student model. ETS leverages VFMs as the expert model in a free lunch manner, thus avoiding significant additional training costs. Extensive experiments exhibit that our method outperforms the existing SOTA methods on three benchmarks.Our code is available in the supplementary materials.