Tag: Large Multimodal Models
All the articles with the tag "Large Multimodal Models".
-
Task-Oriented Semantic Communication in Large Multimodal Models-based Vehicle Networks
This paper proposes a task-oriented semantic communication framework for LMM-based vehicle AI, using LLaVA with Semantic Matching for efficient image slicing and Fusion Attention-based power allocation to prioritize critical data transmission, achieving significant accuracy improvements (up to 33.1% at low SNR) in traffic VQA tasks.