1. DatasetCreation and Annotation: Create or curate high-quality point cloud datasets
specifically designed for segmentation research. This involves collecting or assembling
diverse datasets, preprocessing datasets, and accurately annotating point-level
segmentation masks to serve as ground truth for training and evaluation.
2. AlgorithmDevelopment: Develop novel deep learning architectures, models, or
algorithms for the segmentation tasks. This may involve exploring new network
architectures, loss functions, or training strategies that can improve segmentation
accuracy, handle complex scenarios, or address specific challenges in point cloud
segmentation.
Accessibility Controls