Identity-aware Facial Expression Recognition in Compressed Video

Liu, Xiaofeng; Jin, Linghao; Han, Xu; Lu, Jun; You, Jane; Kong, Lingsheng

doi:10.48550/arxiv.2101.00317

Cited by 1 publication

(1 citation statement)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with over one million images of the ImageNet dataset, the collection of large-scale medical data is challenging for clinical applications (Liu et al 2020b(Liu et al , 2018dHe et al 2020a). To counter this, UDA has gradually become popular (Zou et al 2019;Liu et al 2020cLiu et al , 2021b, which aims to match covariate shift (i.e., only p(x) shift). Discrepancy-based methods (Long et al 2015), such as minimizing MMD, address the dataset shift by mitigating specific discrepancies defined on different layers of a shared model between domains.…”

Section: Related Workmentioning

confidence: 99%

Subtype-aware Unsupervised Domain Adaptation for Medical Diagnosis

Liu

et al. 2021

AAAI

Self Cite

View full text Add to dashboard Cite

Recent advances in unsupervised domain adaptation (UDA) show that transferable prototypical learning presents a powerful means for class conditional alignment, which encourages the closeness of cross-domain class centroids. However, the cross-domain inner-class compactness and the underlying fine-grained subtype structure remained largely underexplored. In this work, we propose to adaptively carry out the fine-grained subtype-aware alignment by explicitly enforcing the class-wise separation and subtype-wise compactness with intermediate pseudo labels. Our key insight is that the unlabeled subtypes of a class can be divergent to one another with different conditional and label shifts, while inheriting the local proximity within a subtype. The cases with or without the prior information on subtype numbers are investigated to discover the underlying subtype structure in an online fashion. The proposed subtype-aware dynamic UDA achieves promising results on a medical diagnosis task.

show abstract