Introduces M3C, a machine-generated multimodal multi-session, multi-party conversation dataset combining images and audio, plus a Qwen2-VL-based model with multimodal memory retrieval.
Also, it learns to establish memory links between ses- sion memory and observed images/audio, as well as between session memory and chat his- tory
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Enabling Chatbots with Eyes and Ears: An Immersive Multimodal Conversation System for Dynamic Interactions
Introduces M3C, a machine-generated multimodal multi-session, multi-party conversation dataset combining images and audio, plus a Qwen2-VL-based model with multimodal memory retrieval.