Multi-agent DQN with shared knowledge among K agents detects multiple landmarks, reducing error by 50% versus separate training.
In: 2012 IEEE 36th Annual Computer Software and Applications Conference
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multiple Landmark Detection using Multi-Agent Reinforcement Learning
Multi-agent DQN with shared knowledge among K agents detects multiple landmarks, reducing error by 50% versus separate training.