SAT: 2D Semantics Assisted Training for 3D Visual Grounding.
Zhengyuan YangSongyang ZhangLiwei WangJiebo LuoPublished in: ICCV (2021)
Keyphrases
- search algorithm
- semantically relevant
- semantic information
- online learning
- high level semantics
- visual information
- hidden markov models
- supervised learning
- logic programming
- training samples
- training examples
- visual perception
- sat solvers
- training process
- training algorithm
- high level
- formal semantics
- training phase
- training set
- visual features
- active learning
- low level