| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
| Presentation Title |
Self-supervised vision transformers for anomaly detection in 3D printing |
| Author(s) |
Bowen Zheng, Xingquan Wang, Zeqing Jin, Grace Gu |
| On-Site Speaker (Planned) |
Bowen Zheng |
| Abstract Scope |
Fused filament fabrication (FFF) faces challenges in printing quality such as under- and over-extrusion, which can degrade printed parts in mechanical properties and surface quality. Previous research has used computer vision to detect printing defects. However, these techniques are mostly based on supervised learning, and require costly manual data labeling. In this work, we combine self-supervised learning and transformers to detect anomalies in FFF 3D printing in an annotation-efficient manner. Using tens of thousands of unlabeled frames recorded by a camera as training data, our model can segment printed area and background and discover printing defects with no supervision or segmentation-targeted objective. Additionally, the model is intrinsically interpretable, thus contributing to a higher-level of image understanding. Our work presents a data-driven anomaly detection technique less dependent on labeled data, which may be important for domains where annotated images are scarce such as in additive manufacturing and medical imaging. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Additive Manufacturing, |