Figure 7From: Performance Analysis of Bit-Width Reduced Floating-Point Arithmetic Units in FPGAs: A Case Study of Neural Network-Based Face Detector Error model for general neural network. Back to article page