A new finite-state vector quantizer with optimized state space and derailment-free operation.
Xavier GinestaSeung P. KimPublished in: IEEE Trans. Image Process. (1996)
Keyphrases
- finite state
- vector quantizer
- state space
- markov decision processes
- markov chain
- optimal policy
- continuous time markov process
- subband image coding
- reinforcement learning
- heuristic search
- bit allocation
- reinforcement learning algorithms
- entropy constrained
- average cost
- dynamic programming
- action sets
- partially observable markov decision processes
- model checking
- tree automata
- infinite horizon
- partially observable
- markov decision process
- policy iteration
- decision problems
- search space
- variable length
- planning problems
- continuous state
- dynamical systems
- vector quantization
- action space
- residual vector quantization