Few-Shot Image-to-Semantics Translation for Policy Transfer in Reinforcement Learning.
Rei SatoKazuto FukuchiJun SakumaYouhei AkimotoPublished in: IJCNN (2022)
Keyphrases
- reinforcement learning
- image data
- image features
- input image
- image classification
- image representation
- image content
- image retrieval
- image analysis
- image segmentation
- multiscale
- image collections
- image pixels
- high resolution
- region of interest
- single image
- segmentation method
- pixel values
- machine translation
- segmentation algorithm
- transfer learning
- optimal policy
- video data
- feature points
- edge detection
- similarity measure
- partially observable
- markov decision problems
- policy search
- action selection
- model free
- video content
- image regions
- detection algorithm
- low level