Task-Oriented Multi-User Semantic Communications for VQA

Semantic communications focus on the transmission of semantic features. In this letter, we consider a task-oriented multi-user communication system for multimodal data transmission. Particularly, partial users transmit images while others texts to inquiry information about images. To exploit correlation among from multiple users, propose deep neural network enabled system, named MU-DeepSC, execute visual question answering (VQA) task as an example. Specifically, transceiver MU-DeepSC is designed and optimized jointly capture features correlated Simulation results demonstrate that proposed more robust channel variations than traditional systems, especially in low signal-to-noise (SNR) regime.