Fig. 16From: A survey on GAN acceleration using memory compression techniquesQuantization decisionsBack to article page