Content area

Abstract

Geometric representation learning can address challenges that were previously difficult for data-driven methods due to data scarcity. Geometry data scarcity can be mitigated through grammar- based modeling or modality conversion, while label scarcity can be tackled in two ways. First, when indirect, easily accessible labels are available, weakly supervised learning allows for the extraction of high-level design features. Second, in the complete absence of labels, inter- modality geometric pretraining improves design quantity estimation in few-shot scenarios. This approach is effective for tasks involving scalar values, temporal histories, and scalar fields. Furthermore, customized training strategies can be tailored to capture and process domain-specific geometries, such as thin shells and geometries with fine-scale details.

Details

1010268
Business indexing term
Title
Geometric Representation Learning for Accelerated Design Analysis in Data-Scarce Environments
Author
Number of pages
148
Publication year
2025
Degree date
2025
School code
0041
Source
DAI-A 87/1(E), Dissertation Abstracts International
ISBN
9798288853074
Committee member
Oancea, Victor; Wang, Liwei
University/institution
Carnegie Mellon University
Department
Mechanical Engineering
University location
United States -- Pennsylvania
Degree
Ph.D.
Source type
Dissertation or Thesis
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
32001811
ProQuest document ID
3231830826
Document URL
https://www.proquest.com/dissertations-theses/geometric-representation-learning-accelerated/docview/3231830826/se-2?accountid=208611
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Database
ProQuest One Academic