SimCS: Simulation for Domain Incremental Online Continual Segmentation

被引:0
|
作者
Alfarra, Motasem [1 ,2 ]
Cai, Zhipeng [1 ]
Bibi, Adel [3 ]
Ghanem, Bernard [2 ]
Mueller, Matthias [1 ]
机构
[1] Intel Labs, Thuwal, Saudi Arabia
[2] King Abdullah Univ Sci & Technol KAUST, Thuwal, Saudi Arabia
[3] Univ Oxford, Oxford, England
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
Continual Learning is a step towards lifelong intelligence where models continuously learn from recently collected data without forgetting previous knowledge. Existing continual learning approaches mostly focus on image classification in the class-incremental setup with clear task boundaries and unlimited computational budget. This work explores the problem of Online Domain-Incremental Continual Segmentation (ODICS), where the model is continually trained over batches of densely labeled images from different domains, with limited computation and no information about the task boundaries. ODICS arises in many practical applications. In autonomous driving, this may correspond to the realistic scenario of training a segmentation model over time on a sequence of cities. We analyze several existing continual learning methods and show that they perform poorly in this setting despite working well in class-incremental segmentation. We propose SimCS, a parameter-free method complementary to existing ones that uses simulated data to regularize continual learning. Experiments show that SimCS provides consistent improvements when combined with different CL methods.
引用
收藏
页码:10795 / 10803
页数:9
相关论文
共 50 条
  • [1] Adaptive Online Domain Incremental Continual Learning
    Gunasekara, Nuwan
    Gomes, Heitor
    Bifet, Albert
    Pfahringer, Bernhard
    ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING - ICANN 2022, PT I, 2022, 13529 : 491 - 502
  • [2] Adaptive Neural Networks for Online Domain Incremental Continual Learning
    Gunasekara, Nuwan
    Gomes, Heitor
    Bifet, Albert
    Pfahringer, Bernhard
    DISCOVERY SCIENCE (DS 2022), 2022, 13601 : 89 - 103
  • [3] Continual Learning for Class- and Domain-Incremental Semantic Segmentation
    Kalb, Tobias
    Roschani, Masoud
    Ruf, Miriam
    Beyerer, Juergen
    2021 32ND IEEE INTELLIGENT VEHICLES SYMPOSIUM (IV), 2021, : 1345 - 1351
  • [4] Continual Unsupervised Domain Adaptation for Semantic Segmentation by Online Frequency Domain Style Transfer
    Termoehlen, Jan-Aike
    Klingner, Marvin
    Brettin, Leon J.
    Schmidt, Nico M.
    Fingscheidt, Tim
    2021 IEEE INTELLIGENT TRANSPORTATION SYSTEMS CONFERENCE (ITSC), 2021, : 2881 - 2888
  • [5] Continual Class Incremental Learning for CT Thoracic Segmentation
    Elskhawy, Abdelrahman
    Lisowska, Aneta
    Keicher, Matthias
    Henry, Joseph
    Thomson, Paul
    Navab, Nassir
    DOMAIN ADAPTATION AND REPRESENTATION TRANSFER, AND DISTRIBUTED AND COLLABORATIVE LEARNING, DART 2020, DCL 2020, 2020, 12444 : 106 - 116
  • [6] Attacks on Continual Semantic Segmentation by Perturbing Incremental Samples
    Yu, Zhidong
    Yang, Wei
    Xie, Xike
    Shi, Zhenbo
    THIRTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 38 NO 7, 2024, : 6844 - 6852
  • [7] Continual Attentive Fusion for Incremental Learning in Semantic Segmentation
    Yang, Guanglei
    Fini, Enrico
    Xu, Dan
    Rota, Paolo
    Ding, Mingli
    Tang, Hao
    Alameda-Pineda, Xavier
    Ricci, Elisa
    IEEE TRANSACTIONS ON MULTIMEDIA, 2023, 25 : 3841 - 3854
  • [8] WEAKLY-SUPERVISED CONTINUAL LEARNING FOR CLASS-INCREMENTAL SEGMENTATION
    Lenczner, Gaston
    Chan-Hon-Tong, Adrien
    Luminari, Nicola
    Le Saux, Bertrand
    2022 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS 2022), 2022, : 4843 - 4846
  • [9] Online Class-Incremental Continual Learning with Adversarial Shapley Value
    Shim, Dongsub
    Mai, Zheda
    Jeong, Jihwan
    Sanner, Scott
    Kim, Hyunwoo
    Jang, Jongseong
    THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 9630 - 9638
  • [10] Continual Domain Incremental Learning for Privacy-Aware Digital Pathology
    Kumari, Pratibha
    Reisenbuechler, Daniel
    Luttner, Lucas
    Schaadt, Nadine S.
    Feuerhake, Friedrich
    Merhof, Dorit
    MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2024, PT XII, 2024, 15012 : 34 - 44