From Easy to Hard: Learning Curricular Shape-aware Features for Robust Panoptic Scene Graph Generation.
Hanrong ShiLin LiJun XiaoYueting ZhuangLong ChenPublished in: CoRR (2024)
Keyphrases
- rotational invariant
- invariant features
- learning process
- learning algorithm
- image sequences
- reinforcement learning
- prior knowledge
- image features
- input image
- visual appearance
- technology enhanced learning
- depth map
- point features
- object features
- learning opportunities
- feature descriptors
- graph structure
- learning analytics
- feature set
- feature space
- video sequences
- three dimensional