Food image generation using a large amount of food images with conditional GAN: ramenGAN and recipeGAN
Food image generation using a large amount of food images with conditional GAN: ramenGAN and recipeGAN
复制标题
DOI：
10.1145/3230519.3230598
复制
发表时间：
2018-07
期刊：
影响因子：
--
通讯作者：
Yoshifumi Ito;Wataru Shimoda;Keiji Yanai
中科院分区：
文献类型：
--
作者：
Yoshifumi Ito;Wataru Shimoda;Keiji Yanai
近年来，基于深度卷积神经网络的图像生成技术得到了广泛的研究。在本文中，我们描述了基于CNN的图像生成的食物图像。特别是，我们专注于使用具有大规模数据集的条件生成对抗网络（cGAN）生成图像。在实验中，我们用“拉面”图像数据集和食谱图像数据集训练cGAN。对于“拉面“GAN，我们添加了一个盘子形状，使生成的图像中的盘子形状更圆。对于“食谱“GAN，我们从烹饪原料中生成菜肴图像，并尝试使用食谱数据库中生成的图像进行基于图像的食谱搜索。
Recently, image generation by Deep Convolutional Neural Network has been studied widely by many researchers. In this paper, we describe CNN-based image generation on food images. Especially, we focus on image generation using conditional Generative Adversarial Network (cGAN) with a large-scale dataset. In the experiments, we trained cGAN with a "ramen" image dataset and a recipe image dataset. For "ramen"GAN, we added a dish plate discriminator to make the shape of dishes rounder in generated images. For "recipe"GAN, we generated dish images from cooking ingredients, and tried image-based recipe search with generated images for the recipe database.