Resource-efficient VLSI Architecture of Softmax Activation Function for Real-time Inference in Deep Learning Applications | IEEE Conference Publication | IEEE Xplore