Segment Any Object Model (SAOM): Real-to-Simulation Fine-Tuning Strategy for Multi-Class Multi-Instance Segmentation.
Mariia KhanYue QiuYuren CongJumana Abu-KhalafDavid SuterBodo RosenhahnPublished in: CoRR (2024)
Keyphrases
- multi class
- multi instance
- object model
- fine tuning
- binary classification
- object segmentation
- support vector machine
- data model
- object detection
- segmentation algorithm
- d objects
- level set
- pairwise
- feature selection
- database
- segmentation method
- cost sensitive
- image segmentation
- real valued
- shape prior
- multiple instance learning
- semi supervised
- semi supervised learning
- prediction accuracy
- multi label
- machine learning