2022 12th International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER) 2022
DOI: 10.1109/cyber55403.2022.9907146
|View full text |Cite
|
Sign up to set email alerts
|

Semi-Supervised Confidence-Level-based Contrastive Discrimination for Class-Imbalanced Semantic Segmentation

Abstract: To overcome the data hungry challenge, we have proposed a semi-supervised contrastive learning framework for the task of class-imbalanced semantic segmentation. First and foremost, to make the model operate in a semi-supervised manner, we proposed the confidence-level-based contrastive learning to achieve instance discrimination in an explicit manner, and make the low-confidence low-quality features align with the high-confidence counterparts. Moreover, to tackle the problem of class imbalance in crack segment… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
6
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
4
3

Relationship

1
6

Authors

Journals

citations
Cited by 8 publications
(6 citation statements)
references
References 21 publications
0
6
0
Order By: Relevance
“…Our network framework and innovatively designed modules to merge over-divided clusters to provide pseudo labels are illustrated in Subsection B and C for segmentation and detection respectively. Also, the LiDAR-based approaches are of significance to many industrial applications such as UAV, UGV as well as service robotic navigation as well as inspections [1]- [3] and robotic enhanced large-scale localization in the diverse complex environments [4]- [12], and large-scale robotic semantic scene parsing [13]- [15], as well as robotic control as well as robotic manipulation applications [16]- [21], etc.…”
Section: Proposed Methodologymentioning
confidence: 99%
“…Our network framework and innovatively designed modules to merge over-divided clusters to provide pseudo labels are illustrated in Subsection B and C for segmentation and detection respectively. Also, the LiDAR-based approaches are of significance to many industrial applications such as UAV, UGV as well as service robotic navigation as well as inspections [1]- [3] and robotic enhanced large-scale localization in the diverse complex environments [4]- [12], and large-scale robotic semantic scene parsing [13]- [15], as well as robotic control as well as robotic manipulation applications [16]- [21], etc.…”
Section: Proposed Methodologymentioning
confidence: 99%
“…Also, it has recently be studied that the image can bridge the big semantic gap between the modalities of text and 3D shapes. Also, the LiDAR-based approaches are of significance to many industrial applications such as UAV/robotics inspections [31,32,36] and robotic enhanced large-scale localization in the diverse complex environments [26, 28-30, 35, 41, 42, 46, 49], and large-scale robotic scene parsing [37][38][39], as well as robotic control as well as manipulation applications [27,33,34,43,45,50], etc. Differently, we make the first attempt in traditional and learnt 3D descriptor guided weakly supervised point cloud segmentation.…”
Section: Learning-based Point Clouds Understanding Methodsmentioning
confidence: 99%
“…The weakly supervised approaches for point cloud understanding are effective manners to reduce high annotation burdens [32]. Many preliminary attempts have been tried including labeling a small portion of points [16,24,52,73] or semantic classes [69].…”
Section: Weakly Supervised Methods For Point Clouds Understandingmentioning
confidence: 99%
“…Most importantly, a significant module for Unmanned Aerial Vehicle (UAV) intelligent inspection systems is to develop computer vision algorithms for processing images captured and detecting cracks and structural damages. The visual sensors have been emerging as novel techniques for modeling the environment [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], and has been widely adopted for robotic applications. Various visual sensors and Light Detection And Ranging Sensors (LiDAR) sensors have been widely deployed in various robotics platforms such as automatically navigated aerial robots and ground robots [16], [17], [18], [19], [20], [21].…”
Section: B the Overall System For Uav Inspectionsmentioning
confidence: 99%
“…Various visual sensors and Light Detection And Ranging Sensors (LiDAR) sensors have been widely deployed in various robotics platforms such as automatically navigated aerial robots and ground robots [16], [17], [18], [19], [20], [21]. For example, the SLAM is widely adopted for robot localization, mapping, and navigation applications [11], [22], [23], [12], [24], [25], [26], [27], [28], [29], [30], [31], [22].…”
Section: B the Overall System For Uav Inspectionsmentioning
confidence: 99%