Modeling Motion with Multi-Modal Features for Text-Based Video Segmentation.
Wangbo ZhaoKai WangXiangxiang ChuFuzhao XueXinchao WangYang YouPublished in: CVPR (2022)
Keyphrases
- multi modal
- video segmentation
- motion cues
- humanoid robot
- video sequences
- multi modality
- semantic concepts
- high dimensional
- low level
- image features
- audio visual
- feature vectors
- video analysis
- video frames
- shot boundary detection
- cross modal
- computer vision
- feature space
- multimedia
- multiple modalities
- uni modal
- motion features
- image annotation
- video streams
- segmentation method
- semantic information
- mutual information