A Deep Reinforcement Learning Approach To Audio-Based Navigation In A Multi-Speaker Environment.
Petros GiannakopoulosAggelos PikrakisYannis CotronisPublished in: ICASSP (2021)
Keyphrases
- reinforcement learning
- indoor environments
- navigation tasks
- speaker identification
- mobile robot
- audio visual
- real time
- indoor navigation
- speech recognition
- automatic transcription
- speaker recognition
- dynamic environments
- autonomous agents
- multimedia
- audio stream
- multi agent environments
- multi agent systems
- dynamic programming
- navigation systems
- information space
- signal processing
- function approximation
- optimal policy
- complex environments
- reinforcement learning algorithms
- noisy environments
- autonomous vehicles
- topological map
- learning process
- service robots
- markov decision processes
- visual information
- path planning
- multi agent