Few-Shot Image-to-Semantics Translation for Policy Transfer in Reinforcement Learning.
Rei SatoKazuto FukuchiJun SakumaYouhei AkimotoPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- image features
- image pixels
- image content
- single image
- image data
- image classification
- low level
- multiscale
- input image
- image retrieval
- image segmentation
- image analysis
- image collections
- segmentation method
- key frames
- learning algorithm
- image representation
- edge detection
- pixel values
- similarity measure
- test images
- region of interest
- temporal difference
- reward function
- markov decision process
- spatial information
- image regions
- optimal policy
- state space
- dynamic programming
- high resolution
- image processing