（五十三）：Deep Visual-Semantic Alignments for Generating Image Descriptions

（五十三）：Deep Visual-Semantic Alignments for Generating Image Descriptions

• Abstract
• 1. Introduction
• 2 RELATED WORK
• 3 OUR MODEL
• • 3.1 Learning to Align Visual and Language Data 学习对齐视觉和语言数据
  • 3.1.1 Representing Images
  • 3.1.2 Representing Sentences
  • 3.1.3 Alignment Objective 客观对齐
  • 3.1.4 Decoding Text Snippet Alignments to Images解码文本片段对齐到图像
  • 3.2 Multimodal Recurrent Neural Network for Generating Descrip