A Pixel-Level Meta-Learner for Weakly Supervised Few-Shot Semantic Segmentation

被引:6
作者
Lee, Yuan-Hao [1 ]
Yang, Fu-En
Wang, Yu-Chiang Frank
机构
[1] Natl Taiwan Univ, Grad Inst Commun Engn, Taipei, Taiwan
来源
2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022) | 2022年
关键词
D O I
10.1109/WACV51458.2022.00167
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Few-shot semantic segmentation addresses the learning task in which only few images with ground truth pixel-level labels are available for the novel classes of interest. One is typically required to collect a large mount of data (i.e., base classes) with such ground truth information, followed by meta-learning strategies to address the above learning task. When only image-level semantic labels can be observed during both training and testing, it is considered as an even more challenging task of weakly supervised few-shot semantic segmentation. To address this problem, we propose a novel meta-learning framework, which predicts pseudo pixel-level segmentation masks from a limited amount of data and their semantic labels. More importantly, our learning scheme further exploits the produced pixel-level information for query image inputs with segmentation guarantees. Thus, our proposed learning model can be viewed as a pixel-level meta-learner. Through extensive experiments on benchmark datasets, we show that our model achieves satisfactory performances under fully supervised settings, yet performs favorably against state-of-the-art methods under weakly supervised settings.
引用
收藏
页码:1607 / 1617
页数:11
相关论文
共 50 条
[41]   Self-Supervised Tuning for Few-Shot Segmentation [J].
Zhu, Kai ;
Zhai, Wei ;
Cao, Yang .
PROCEEDINGS OF THE TWENTY-NINTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2020, :1019-1025
[42]   A pseudo-labeling based weakly supervised segmentation method for few-shot texture images [J].
Han, Yuexing ;
Li, Ruiqi ;
Wang, Bing ;
Ruan, Liheng ;
Chen, Qiaochuan .
EXPERT SYSTEMS WITH APPLICATIONS, 2024, 238
[43]   StarNet: towards Weakly Supervised Few-Shot Object Detection [J].
Karlinsky, Leonid ;
Shtok, Joseph ;
Alfassy, Amit ;
Lichtenstein, Moshe ;
Harary, Sivan ;
Schwartz, Eli ;
Doveh, Sivan ;
Sattigeri, Prasanna ;
Feris, Rogerio ;
Bronstein, Alexander ;
Giryes, Raja .
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 :1743-1753
[44]   Few-Shot Font Generation With Weakly Supervised Localized Representations [J].
Park, Song ;
Chun, Sanghyuk ;
Cha, Junbum ;
Lee, Bado ;
Shim, Hyunjung .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2024, 46 (03) :1479-1495
[45]   Self-support Few-Shot Semantic Segmentation [J].
Fan, Qi ;
Pei, Wenjie ;
Tai, Yu-Wing ;
Tang, Chi-Keung .
COMPUTER VISION, ECCV 2022, PT XIX, 2022, 13679 :701-719
[46]   Few-Shot Semantic Segmentation with Cyclic Memory Network [J].
Xie, Guo-Sen ;
Xiong, Huan ;
Liu, Jie ;
Yao, Yazhou ;
Shao, Ling .
2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, :7273-7282
[47]   Query semantic reconstruction for background in few-shot segmentation [J].
Haoyan Guan ;
Michael Spratling .
The Visual Computer, 2024, 40 (2) :799-810
[48]   Few-Shot Semantic Segmentation via Mask Aggregation [J].
Ao, Wei ;
Zheng, Shunyi ;
Meng, Yan ;
Yang, Yang .
NEURAL PROCESSING LETTERS, 2024, 56 (02)
[49]   Query semantic reconstruction for background in few-shot segmentation [J].
Guan, Haoyan ;
Spratling, Michael .
VISUAL COMPUTER, 2024, 40 (02) :799-810
[50]   Incorporating Depth Information into Few-Shot Semantic Segmentation [J].
Zhang, Yifei ;
Sidibe, Desire ;
Morel, Olivier ;
Meriaudeau, Fabrice .
2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2021, :3582-3588