Tag
This paper proposes a collaborative memory framework for multi-agent vision-language model systems to address distributed perception and improve shared visual context and reasoning consistency.