当前位置: X-MOL 学术IEEE Trans. Pattern Anal. Mach. Intell. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Where and How to Transfer: Knowledge Aggregation-Induced Transferability Perception for Unsupervised Domain Adaptation
IEEE Transactions on Pattern Analysis and Machine Intelligence ( IF 20.8 ) Pub Date : 2021-11-16 , DOI: 10.1109/tpami.2021.3128560
Jiahua Dong , Yang Cong , Gan Sun , Zhen Fang , Zhengming Ding

Unsupervised domain adaptation without accessing expensive annotation processes of target data has achieved remarkable successes in semantic segmentation. However, most existing state-of-the-art methods cannot explore whether semantic representations across domains are transferable or not, which may result in the negative transfer brought by irrelevant knowledge. To tackle this challenge, in this paper, we develop a novel Knowledge Aggregation-induced Transferability Perception (KATP) module for unsupervised domain adaptation, which is a pioneering attempt to distinguish transferable or untransferable knowledge across domains. Specifically, the KATP module is designed to quantify which semantic knowledge across domains is transferable, by incorporating the transferability information propagation from constructed global category-wise prototypes. Based on KATP, we design a novel KATP Adaptation Network (KATPAN) to determine where and how to transfer. The KATPAN contains a transferable appearance translation module TA(⋅)\mathcal {T}_A(\cdot) and a transferable representation augmentation module TR(⋅)\mathcal {T}_R(\cdot), where both modules construct a virtuous circle of performance promotion. TA(⋅)\mathcal {T}_A(\cdot) develops a transferability-aware information bottleneck to highlight where to adapt transferable visual characterizations and modality information; TR(⋅)\mathcal {T}_R(\cdot) explores how to augment transferable representations while abandoning untransferable information, and promotes the translation performance of TA(⋅)\mathcal {T}_A(\cdot) in return. Comprehensive experiments on several representative benchmark datasets and a medical dataset support the state-of-the-art performance of our model.

中文翻译:


迁移地点和方式:知识聚合引发的无监督领域适应的可迁移性感知



无需访问目标数据的昂贵注释过程的无监督域适应在语义分割方面取得了显着的成功。然而,大多数现有的最先进的方法无法探索跨领域的语义表示是否可迁移,这可能会导致不相关知识带来的负迁移。为了应对这一挑战,在本文中,我们开发了一种新颖的知识聚合诱导的可转移性感知(KATP)模块,用于无监督领域适应,这是区分跨领域可转移或不可转移知识的开创性尝试。具体来说,KATP 模块旨在通过合并来自构建的全局类别原型的可转移性信息传播来量化哪些跨领域的语义知识是可转移的。基于KATP,我们设计了一种新颖的KATP适应网络(KATPAN)来确定转移的地点和方式。 KATPAN 包含一个可转移的外观翻译模块 TA(⋅)\mathcal {T}_A(\cdot) 和一个可转移的表示增强模块 TR(⋅)\mathcal {T}_R(\cdot),其中两个模块构建了一个良性循环的绩效提升。 TA(⋅)\mathcal {T}_A(\cdot) 开发了一个可传递性感知信息瓶颈,以突出在何处适应可传递的视觉特征和模态信息; TR(⋅)\mathcal {T}_R(\cdot) 探索如何在放弃不可传递信息的同时增强可转移表示,并反过来提高 TA(⋅)\mathcal {T}_A(\cdot) 的翻译性能。对几个代表性基准数据集和医学数据集的综合实验支持我们模型的最先进的性能。
更新日期:2021-11-16
down
wechat
bug