Bayesian Semantic Instance Segmentation in Open Set World

被引:17
|
作者
Pham, Trung [1 ]
Kumar, B. G. Vijay [1 ]
Do, Thanh-Toan [1 ]
Carneiro, Gustavo [1 ]
Reid, Ian [1 ]
机构
[1] Univ Adelaide, Sch Comp Sci, Adelaide, SA, Australia
来源
COMPUTER VISION - ECCV 2018, PT X | 2018年 / 11214卷
基金
澳大利亚研究理事会;
关键词
Instance segmentation; Open-set conditions; IMAGE SEGMENTATION;
D O I
10.1007/978-3-030-01249-6_1
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper addresses the semantic instance segmentation task in the open-set conditions, where input images can contain known and unknown object classes. The training process of existing semantic instance segmentation methods requires annotation masks for all object instances, which is expensive to acquire or even infeasible in some realistic scenarios, where the number of categories may increase boundlessly. In this paper, we present a novel open-set semantic instance segmentation approach capable of segmenting all known and unknown object classes in images, based on the output of an object detector trained on known object classes. We formulate the problem using a Bayesian framework, where the posterior distribution is approximated with a simulated annealing optimization equipped with an efficient image partition sampler. We show empirically that our method is competitive with state-of-the-art supervised methods on known classes, but also performs well on unknown classes when compared with unsupervised methods.
引用
收藏
页码:3 / 18
页数:16
相关论文
共 50 条
  • [1] Open-set marine object instance segmentation with prototype learning
    Hu, Xing
    Li, Panlong
    Karimi, Hamid Reza
    Jiang, Linhua
    Zhang, Dawei
    SIGNAL IMAGE AND VIDEO PROCESSING, 2024, 18 (8-9) : 6055 - 6062
  • [2] Bridging the Gap Between Semantic Segmentation and Instance Segmentation
    Yin, Chengxiang
    Tang, Jian
    Yuan, Tongtong
    Xu, Zhiyuan
    Wang, Yanzhi
    IEEE TRANSACTIONS ON MULTIMEDIA, 2022, 24 : 4183 - 4196
  • [3] Open Set Semantic Segmentation for Multitemporal Crop Recognition
    Chamorro Martinez, Jorge A.
    Oliveira, Hugo
    dos Santos, Jefersson A.
    Feitosa, Raul Queiroz
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2022, 19
  • [4] Combine Supervised Edge and Semantic Supplement for Instance Segmentation
    Yang, Yakun
    Luo, Wenjie
    Tian, Xuedong
    IEEE ACCESS, 2022, 10 : 89852 - 89860
  • [5] Box-Supervised Instance Segmentation with Level Set Evolution
    Li, Wentong
    Liu, Wenyu
    Zhu, Jianke
    Cui, Miaomiao
    Hua, Xian-Sheng
    Zhang, Lei
    COMPUTER VISION, ECCV 2022, PT XXIX, 2022, 13689 : 1 - 18
  • [6] Indoor Instance-Aware Semantic Mapping Using Instance Segmentation
    Jiang, Yinpeng
    Ma, Xudong
    Fang, Fang
    Kang, Xuewen
    PROCEEDINGS OF THE 33RD CHINESE CONTROL AND DECISION CONFERENCE (CCDC 2021), 2021, : 3549 - 3554
  • [7] Semantic versus instance segmentation in microscopic algae detection
    Ruiz-Santaquiteria, Jesus
    Bueno, Gloria
    Deniz, Oscar
    Vallez, Noelia
    Cristobal, Gabriel
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2020, 87 (87)
  • [8] Instance Segmentation based Semantic Matting for Compositing Applications
    Hu, Guanqing
    Clark, James J.
    2019 16TH CONFERENCE ON COMPUTER AND ROBOT VISION (CRV 2019), 2019, : 135 - 142
  • [9] Semantic Attention and Scale Complementary Network for Instance Segmentation in Remote Sensing Images
    Zhang, Tianyang
    Zhang, Xiangrong
    Zhu, Peng
    Tang, Xu
    Li, Chen
    Jiao, Licheng
    Zhou, Huiyu
    IEEE TRANSACTIONS ON CYBERNETICS, 2022, 52 (10) : 10999 - 11013
  • [10] A multiple instance learning based framework for semantic image segmentation
    Gondra, Iker
    Xu, Tao
    MULTIMEDIA TOOLS AND APPLICATIONS, 2010, 48 (02) : 339 - 365