基于多视图跨模态特征融合的图像描述生成

张乃洲,赵云超,曹薇,张啸剑

Image captioning generation based on multiple-view cross-modal feature fusion

Naizhou ZHANG,Yunchao ZHAO,Wei CAO,Xiaojian ZHANG

表 2 在MSCOCO测试数据集上与其他先进模型在集成模型上的性能比较

Tab.2 Comparison with other state-of-the-art model on MSCOCO test dataset in ensemble-model setting %

模型 BLEU-1 BLEU-4 METEOR ROUGE-L CIDEr SPICE SCST[6] — 35.4 27.1 56.6 117.5 — AoANet[9] 81.6 40.2 29.3 59.4 132.0 22.8 X-Transformer[10] 81.7 40.7 29.9 59.7 135.3 23.8 M2Transformer[11] 82.0 40.5 29.7 59.5 134.5 23.5 GET[12] 82.1 40.6 29.8 59.6 135.1 23.8 DLCT[14] 82.2 40.8 29.9 59.8 137.5 23.3 PureT[21] 83.4 42.1 30.4 60.8 141.0 24.3 MVCMFAF (本文) 83.5 42.7 30.6 61.1 142.3 24.5