All data are split into train/val/test sets while the clinical dataset serves as a test set for all methods. In addition to synthetic data, real data of different persons are recorded in front of a green screen and in a clinical environment. Likewise, the aim of the presented methodology is to explore a pipeline for the generation of synthetic data for the medical field, so that further research questions from the intervention space can be explored. Both methods (SCANS, CAD) are incorporated into a domain randomization environment called NDDS and a Structured Domain Randomization environment implemented in Unity, based on, to generate synthetic training data. The comparison is performed using the example of medical clothing object detection. This work presents a comparison in terms of detection accuracy and generalizability of different methods for synthetic clothing generation using either 3D clothing scans (SCANS) or designed CAD clothing (CAD) with SMPL models.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |