Fig. 7From: Spatial attention and quantization-based contrastive learning framework for mmWave massive MIMO beam trainingAverage success rate performance of SE-QPTNet, QPTNet, QP-DNN, and LSTM attention [30] with different prior label informationBack to article page