基于多融合模型的图像语义描述研究

Image Caption Based on Multi-fusion Model

图像语义描述可以自动生成图像的自然语言描述,对场景理解具有重要意义。本文主要针对图像语义描述的特征学习和语义学习等进行改进,提出一种新的多融合模型。实验结果表明,本文提出的模型有较好的描述效果,但模型在训练时时间过长,有待改进。

Image semantic description can automatically generate natural language description of images,which is of great significance to scene understanding.In this paper,we proposed a new multi-fusion model for feature learning and semantic learning of image semantic description.The experimental results show that the model proposed in this paper has good descriptive effect,but the training time of the model is too long and needs to be improved.