Robustness via Uncertainty-aware Cycle Consistency

被引:0
|
作者
Upadhyay, Uddeshya [1 ]
Chen, Yanbei [1 ]
Akata, Zeynep [1 ,2 ]
机构
[1] Univ Tubingen, Tubingen, Germany
[2] Max Planck Inst Intelligent Syst, Stuttgart, Germany
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Unpaired image-to-image translation refers to learning inter-image-domain mapping without corresponding image pairs. Existing methods learn deterministic mappings without explicitly modelling the robustness to outliers or predictive uncertainty, leading to performance degradation when encountering unseen perturbations at test time. To address this, we propose a novel probabilistic method based on Uncertainty-aware Generalized Adaptive Cycle Consistency (UGAC), which models the per-pixel residual by generalized Gaussian distribution, capable of modelling heavy-tailed distributions. We compare our model with a wide variety of state-of-the-art methods on various challenging tasks including unpaired image translation of natural images, using standard datasets, spanning autonomous driving, maps, facades, and also in medical imaging domain consisting of MRI. Experimental results demonstrate that our method exhibits stronger robustness towards unseen perturbations in test data. Code is released here: https://github.com/ExplainableML/UncertaintyAwareCycleConsistency.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] Robust Tracking via Uncertainty-Aware Semantic Consistency
    Ma, Jie
    Lan, Xiangyuan
    Zhong, Bineng
    Li, Guorong
    Tang, Zhenjun
    Li, Xianxian
    Ji, Rongrong
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2023, 33 (04) : 1740 - 1751
  • [2] Uncertainty-Aware Robustness Assessment of Industrial Elevator Systems
    Han, Liping
    Ali, Shaukat
    Yue, Tao
    Arrieta, Aitor
    Arratibel, Maite
    ACM TRANSACTIONS ON SOFTWARE ENGINEERING AND METHODOLOGY, 2023, 32 (04)
  • [3] Uncertainty-aware Consistency Learning for Cold-Start Item Recommendation
    Liu, Taichi
    Gao, Chen
    Wang, Zhenyu
    Li, Dong
    Hao, Jianye
    Jin, Depeng
    Li, Yong
    PROCEEDINGS OF THE 46TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL, SIGIR 2023, 2023, : 2466 - 2470
  • [4] Uncertainty-aware consistency regularization for cross-domain semantic segmentation
    Zhou, Qianyu
    Feng, Zhengyang
    Gu, Qiqi
    Cheng, Guangliang
    Lu, Xuequan
    Shi, Jianping
    Ma, Lizhuang
    Computer Vision and Image Understanding, 2022, 221
  • [5] Uncertainty-aware consistency regularization for cross-domain semantic segmentation
    Zhou, Qianyu
    Feng, Zhengyang
    Gu, Qiqi
    Cheng, Guangliang
    Lu, Xuequan
    Shi, Jianping
    Ma, Lizhuang
    COMPUTER VISION AND IMAGE UNDERSTANDING, 2022, 221
  • [6] Uncertainty-Aware Prognosis via Deep Gaussian Process
    Biggio, Luca
    Wieland, Alexander
    Chao, Manuel Arias
    Kastanis, Iason
    Fink, Olga
    IEEE ACCESS, 2021, 9 : 123517 - 123527
  • [7] Uncertainty-aware consistency learning for semi-supervised medical image segmentation
    Dong, Min
    Yang, Ating
    Wang, Zhenhang
    Li, Dezhen
    Yang, Jing
    Zhao, Rongchang
    KNOWLEDGE-BASED SYSTEMS, 2025, 309
  • [8] Uncertainty-Aware Sequence Labeling
    Ye, Jiacheng
    Zhou, Xiang
    Zheng, Xiaoqing
    Gui, Tao
    Zhang, Qi
    IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2022, 30 : 1775 - 1788
  • [9] Uncertainty-Aware Multidimensional Scaling
    Hagele D.
    Krake T.
    Weiskopf D.
    IEEE Transactions on Visualization and Computer Graphics, 2023, 29 (01) : 23 - 32
  • [10] Uncertainty-Aware Pedestrian Crossing Prediction via Reinforcement Learning
    Dai, Siyang
    Liu, Jun
    Cheung, Ngai-Man
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2024, 34 (10) : 9540 - 9549