zhoujunr1的博客

1. Model Overview1.1 什么是weakly supervised learninghttps://stackoverflow.com/questions/18944805/what-is-weakly-supervised-learning-bootstrappingIn short: In weakly supervised learning, you use a limited

Show and Tell Lessons learned from the 2015 MSCOCO Image Captioning Challenge论文及tensorflow源码解读

论文解读Network-in-networkR-CNNMultivation and High Level ConsiderationsArchitectural Details源码分析函数源码Going deeper with convolutions Christian Szegedy, Wei Liu, Yangqing Jia, Pierre...

Source codebuild_modelbuild_image_embeddingsSource codebuild_modelbuild_image_embeddings在建立了图片和caption的输入后，这部分将图片转换为固定大小的tensor，就像论文提及的，使用已经用很大的数据集训练好的深度网络模型，不改变它的参数，直接用于特征提取。首先将图片丢入inception v3网络中，得

论文解读先稍微翻译下数据集ImageNet 有各种分辨率的图片，我们需要固定大小的。所以，向下取样到 256 * 256。除此之外没有别的预处理。 ImageNet consists of variable-resolution images, while our system requires a constant input dimen- sionality. Therefore, we do