Generalized few-shot 3D point cloud segmentation
Few-Shot 3D Point Cloud Semantic Segmentation (3D-FS) mitigates the issues of insufficient data annotation and emerging novel classes in real-world scenarios, but it totally ignores the performance on base classes. In this paper, we address a more practical task, Generalized Few-Shot 3D Point Clo...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Research |
| Language: | English |
| Published: |
Nanyang Technological University
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/179000 |
| _version_ | 1826114767712944128 |
|---|---|
| author | Yang, Shuqian |
| author2 | Jiang Xudong |
| author_facet | Jiang Xudong Yang, Shuqian |
| author_sort | Yang, Shuqian |
| collection | NTU |
| description | Few-Shot 3D Point Cloud Semantic Segmentation (3D-FS) mitigates the issues of insufficient data annotation and emerging novel classes in real-world scenarios, but it totally ignores the performance on base classes.
In this paper, we address a more practical task, Generalized Few-Shot 3D Point Cloud Semantic Segmentation (3D-GFS), which aims to perform segmentation simultaneously on base classes with adequate samples and novel classes with few samples. Based on the prototypical Base Model, we propose Adaptive Support Enrichment module and Query Aware Representation module to utilize the contextual information of semantic segmentation. The former exploits the co-relationship between base and novel classes in support samples while the latter mines semantic information from query samples. Besides, considering the different embedding spaces, we propose a new training strategy to get a better representation of prototypes. Experiments on S3DIS and ScanNet show that our proposed method outperforms our Base Model and the conventional 3D-FS methods. |
| first_indexed | 2024-10-01T03:44:33Z |
| format | Thesis-Master by Research |
| id | ntu-10356/179000 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:44:33Z |
| publishDate | 2024 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1790002024-08-01T08:11:46Z Generalized few-shot 3D point cloud segmentation Yang, Shuqian Jiang Xudong School of Electrical and Electronic Engineering Rapid-Rich Object Search (ROSE) Lab EXDJiang@ntu.edu.sg Computer and Information Science Engineering 3D point cloud Semantic segmentation Generalized few-shot segmentation Few-Shot 3D Point Cloud Semantic Segmentation (3D-FS) mitigates the issues of insufficient data annotation and emerging novel classes in real-world scenarios, but it totally ignores the performance on base classes. In this paper, we address a more practical task, Generalized Few-Shot 3D Point Cloud Semantic Segmentation (3D-GFS), which aims to perform segmentation simultaneously on base classes with adequate samples and novel classes with few samples. Based on the prototypical Base Model, we propose Adaptive Support Enrichment module and Query Aware Representation module to utilize the contextual information of semantic segmentation. The former exploits the co-relationship between base and novel classes in support samples while the latter mines semantic information from query samples. Besides, considering the different embedding spaces, we propose a new training strategy to get a better representation of prototypes. Experiments on S3DIS and ScanNet show that our proposed method outperforms our Base Model and the conventional 3D-FS methods. Master's degree 2024-07-17T01:12:42Z 2024-07-17T01:12:42Z 2024 Thesis-Master by Research Yang, S. (2024). Generalized few-shot 3D point cloud segmentation. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/179000 https://hdl.handle.net/10356/179000 10.32657/10356/179000 en This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). application/pdf Nanyang Technological University |
| spellingShingle | Computer and Information Science Engineering 3D point cloud Semantic segmentation Generalized few-shot segmentation Yang, Shuqian Generalized few-shot 3D point cloud segmentation |
| title | Generalized few-shot 3D point cloud segmentation |
| title_full | Generalized few-shot 3D point cloud segmentation |
| title_fullStr | Generalized few-shot 3D point cloud segmentation |
| title_full_unstemmed | Generalized few-shot 3D point cloud segmentation |
| title_short | Generalized few-shot 3D point cloud segmentation |
| title_sort | generalized few shot 3d point cloud segmentation |
| topic | Computer and Information Science Engineering 3D point cloud Semantic segmentation Generalized few-shot segmentation |
| url | https://hdl.handle.net/10356/179000 |
| work_keys_str_mv | AT yangshuqian generalizedfewshot3dpointcloudsegmentation |