Morphological Galaxy Classification Using Convolutional Neural Networks on FPGA

Deep learning techniques are empowering many space based applications with good speed and accuracy. The application includes categorizing astronomical data, identifying celestial bodies, tracking their movements and many more. In this paper, we present CNN models for morphological galaxy classification on hardware. CNN models namely SqueezeNet, MobileNet, ResNet and EfficientNet trained on Galaxy10 SDSS Dataset. We implement and analyze the performance of trained deep learning models on pre-built deep learning processor for Xilinx ZCU102 provided by MATLAB’s Deep Learning HDL Toolbox Support Package for Xilinx FPGA and SoC Devices. The performance results are also reported in this work.