Making Video Recognition Models Robust to Common Corruptions With Supervised Contrastive Learning.
Tomu HirataYusuke MukutaTatsuya HaradaPublished in: MMAsia (2021)
Keyphrases
- learning algorithm
- learning process
- supervised learning
- unsupervised learning
- learning models
- prior knowledge
- video data
- recognition rate
- accurate models
- visual recognition
- hidden variables
- machine learning
- online learning
- structured prediction
- weakly supervised
- object recognition
- pattern recognition
- multimedia
- model selection
- active learning
- learning tasks
- video sequences
- learning environment
- learned models
- computer vision
- convolutional neural networks