上一条: Energy-efficient and quality-assured approximate computing framework using a co-training method
下一条: ITT-RNA: Imperfection Tolerable Training for RRAM-Crossbar-Based Deep Neural-Network Accelerator