I am Vishal Thengane, a joint PhD student at the University of Surrey (UK) and the University of Wollongong (Australia), supervised by Dr. Xiatian Zhu and Dr. Salim Bouzerdoum.
My research focuses on 3D computer vision, point cloud understanding, and continual learning, with an emphasis on few-shot and class-incremental learning for 3D scene understanding. I am particularly interested in developing scalable learning systems that can adapt to new classes and environments while preserving previously learned knowledge.
- ContinualCLIP: First work showing that CLIP can be extended to continual learning in a zero-shot setting.
- CLIMB-3D: Continual learning for long-tail point cloud instance segmentation.
- SCOPE: Few-shot class-incremental learning for point cloud segmentation.
- Continual Learning
- Few-Shot Learning
- 3D Computer Vision
- Point Cloud Segmentation
- Foundation Models for 3D Understanding