SeekNet: Improved Human Instance Segmentation via Reinforcement Learning Based Optimized Robot Relocation.
Venkatraman NarayananBala Murali ManogharRama Prashanth RVAniket BeraPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- human robot interaction
- human robot
- robot behavior
- segmentation algorithm
- robot control
- level set
- real robot
- image segmentation
- action selection
- perceptual aliasing
- segmentation method
- mobile robot
- medical images
- human users
- object segmentation
- learning algorithm
- robot navigation
- region growing
- human operators
- human subjects
- autonomous learning
- position and orientation
- autonomous robots
- state space
- robot manipulators
- service robots
- dynamic programming
- humanoid robot
- multi robot
- multiscale
- sensory inputs
- energy function
- graph cuts
- vision system
- segmentation accuracy
- edge detection
- model free
- human beings
- medical imaging
- path planning
- markov decision processes
- function approximation
- shape prior