Pretraining the Vision Transformer Using Self-Supervised Methods for Vision Based Deep Reinforcement Learning.
Manuel GoulãoArlindo L. OliveiraPublished in: ECAI (2023)
Keyphrases
- supervised methods
- reinforcement learning
- vision system
- computer vision
- semantic segmentation
- labeled data
- semi supervised
- weakly supervised
- word sense disambiguation
- unsupervised methods
- labor intensive
- transfer learning
- methods require
- labeled training data
- supervised learning
- object class
- image processing
- learning algorithm
- machine learning
- classification models
- unlabeled data
- state space
- error prone
- data sets