Multi-Modal Video Topic Segmentation with Dual-Contrastive Domain Adaptation.
Linzi XingQuan Hung TranFabian CabaFranck DernoncourtSeunghyun YoonZhaowen WangTrung BuiGiuseppe CareniniPublished in: CoRR (2023)
Keyphrases
- multi modal
- domain adaptation
- topic segmentation
- semantic concepts
- multiple sources
- cross domain
- video data
- semi supervised
- labeled data
- transfer learning
- multimedia
- semi supervised learning
- sentiment classification
- video sequences
- unlabeled data
- target domain
- training data
- video content
- image annotation
- video clips
- document classification
- video retrieval
- co training
- data mining
- active learning
- image classification
- sentiment analysis
- bag of words