Please use this identifier to cite or link to this item:
Title: Tackling background ambiguities in multi-class few-shot point cloud semantic segmentation
Authors: Lai, Lvlong
Chen, Jian
Zhang, Chi
Zhang, Zehong
Lin, Guosheng
Wu, Qingyao
Keywords: Engineering::Computer science and engineering
Issue Date: 2022
Source: Lai, L., Chen, J., Zhang, C., Zhang, Z., Lin, G. & Wu, Q. (2022). Tackling background ambiguities in multi-class few-shot point cloud semantic segmentation. Knowledge-Based Systems, 253, 109508-.
Project: MOE-T2EP20220-0007
Journal: Knowledge-Based Systems
Abstract: Few-shot point cloud semantic segmentation learns to segment novel classes with scarce labeled samples. Within an episode, a novel target class is defined by a few support samples with corresponding binary masks, where only the points of this class are labeled as foreground and others are regarded as background. In the tasks involving multiple target classes, since the meanings of background are diverse for different target classes, background ambiguities appear: Some points labeled as background in one support sample may be of other target classes. It will result in incorrect guidance and damage model's segmentation performance. However, previous methods in the literature do not consider this problem. In this paper, we propose a simple yet effective approach to tackle background ambiguities, which adopts the entropy of predictions on query samples to the training objective function as an additional regularization. Besides, we design a feature transformation operation to reduce the feature differences between support and query samples. With our proposed approach, fine-tuning, a weak baseline method for few-shot segmentation, gains significant performance improvement (e.g., 7.48% and 7.04% in 2-way-1-shot and 3-way-1-shot tasks of S3DIS, respectively) and outperforms current state-of-the-art methods in all the task settings of S3DIS and ScanNet benchmark datasets.
ISSN: 0950-7051
DOI: 10.1016/j.knosys.2022.109508
Rights: © 2022 Elsevier B.V. All rights reserved.
Fulltext Permission: none
Fulltext Availability: No Fulltext
Appears in Collections:SCSE Journal Articles

Citations 50

Updated on Mar 15, 2023

Page view(s)

Updated on Mar 20, 2023

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.