Multi-threaded System Design of A Multi-Precision Deep Learning Accelerator on FPGA with Optimized Memory Usage | IEEE Conference Publication | IEEE Xplore