基于生成对抗网络和坐标注意力机制的文本生成图像算法

李云红,张琪琪,陈锦妮,陈伟重,苏雪平,梁成名

Text-to-image generation algorithm based on generative adversarial network and coordinate attention mechanism

Yunhong LI,Qiqi ZHANG,Jinni CHEN,Weichong CHEN,Xueping SU,Chengming LIANG

表 2 不同方法评价指标的分析表

Tab.2 Analysis table of evaluation indicator for different method

方法 Oxford-102 CUB-200 COCO IS FID IS FID IS FID AttnGAN[7] 3.57 24.65 4.36 55.40 25.85 35.49 DF-GAN[8] 3.80 17.15 4.86 14.81 25.45 28.92 ViewDiff[10] 3.86 16.34 4.96 15.69 25.59 — NoiseCollage[11] 3.95 17.45 5.03 17.32 — 28.32 DM-GAN[18] 3.46 20.55 4.75 16.09 29.81 — StackGAN++[19] 3.26 18.36 4.04 15.58 26.73 27.03 DAE-GAN[20] 3.97 17.76 4.42 — — — DT-GAN[21] — — 4.88 16.35 26.32 40.21 CAT-GAN 4.10 16.76 5.13 14.34 31.81 26.36