PromptonomyViT: Multi-Task Prompt Learning Improves Video Transformers using Synthetic Scene Data.
Roei HerzigOfir AbramovichElad Ben-AvrahamAssaf ArbelleLeonid KarlinskyAriel ShamirTrevor DarrellAmir GlobersonPublished in: CoRR (2022)
Keyphrases
- multi task
- learning tasks
- data sets
- multi task learning
- prior knowledge
- multiple tasks
- supervised learning
- multitask learning
- data analysis
- video sequences
- training data
- learning problems
- learning models
- image segmentation
- learning process
- data points
- sparse learning
- machine learning algorithms
- high dimensional data
- support vector machine