no code implementations • 2 Jan 2024 • Mario Döbler, Florian Marencke, Robert A. Marsden, Bin Yang
In real-world scenarios, test data streams are not always independent and identically distributed (i. i. d.).
1 code implementation • 1 Jun 2023 • Robert A. Marsden, Mario Döbler, Bin Yang
To tackle the problem of universal TTA, we identify and highlight several challenges a self-training based method has to deal with: 1) model bias and the occurrence of trivial solutions when performing entropy minimization on varying sequence lengths with and without multiple domain shifts, 2) loss of generalization which exacerbates the adaptation to multiple domain shifts and the occurrence of catastrophic forgetting, and 3) performance degradation due to shifts in class prior.
1 code implementation • CVPR 2023 • Mario Döbler, Robert A. Marsden, Bin Yang
We demonstrate the effectiveness of our proposed method 'robust mean teacher' (RMT) on the continual and gradual corruption benchmarks CIFAR10C, CIFAR100C, and Imagenet-C. We further consider ImageNet-R and propose a new continual DomainNet-126 benchmark.
1 code implementation • 16 Aug 2022 • Robert A. Marsden, Mario Döbler, Bin Yang
In this work, we address two problems that exist when applying self-training in the setting of test-time adaptation.
no code implementations • 12 Aug 2022 • Robert A. Marsden, Felix Wiewel, Mario Döbler, Yang Yang, Bin Yang
In this work, we focus on UDA and additionally address the case of adapting not only to a single domain, but to a sequence of target domains.
no code implementations • 7 Mar 2022 • George Eskandar, Robert A. Marsden, Pavithran Pandiyan, Mario Döbler, Karim Guirguis, Bin Yang
Integrating different representations from complementary sensing modalities is crucial for robust scene interpretation in autonomous driving.
no code implementations • 5 May 2021 • Robert A. Marsden, Alexander Bartler, Mario Döbler, Bin Yang
To avoid the costly annotation of training data for unseen domains, unsupervised domain adaptation (UDA) attempts to provide efficient knowledge transfer from a labeled source domain to an unlabeled target domain.
Ranked #20 on Synthetic-to-Real Translation on SYNTHIA-to-Cityscapes