- Visual Relationship Detection with Language Priors 2016
visual relationship detection領域的開山之作,檢測和定位圖上的多對object,然後分別對每一對object的交互關係(predicate)進行分類。
- Scene Graph Generation by Iterative Message Passing 2017
利用場景圖對對象和關係進行建模。
- Visual Genome :Connecting Language and Vision Using Crowdsourced Dense Image Annotations 2017
PaStaNet:Toward Human Activity Knowledge Engine 2020
利用part級別的動作狀態(part states)進行行爲理解,
- A Hierarchical Approach for Generating Descriptive Image Paragraphs 2017
這篇文章提出了一種結構化模型去生成圖像段落描述。文章首先利用區域生成網絡 (RegionProposal Network)進行區域檢測,並將檢測到的區域投影到卷積特徵圖中,通過插值並通過全連接層和最大池化生成一個區域特徵向量,該向量作爲句子RNN和詞RNN兩層層次循環網絡(Hierarchical Recurrent Networks)的輸入並生成對圖像的描述段落。
- Deep Reasoning with Knowledge Graph for Social Relationship Understanding 2018
- Detecting and Recognizing Human-Object Interactions 2018
檢測人和物體的交互(interaction)定位和人交互的物體的位置並通過人和物來預測後面的動作
- Object Detection Meets Knowledge Graphs 2017
引入外部知識庫來進行輔助目標檢測。
- The More You Know: Using Knowledge Graphs for Image Classification 2016
應用知識圖譜形式的結構化先驗知識到圖像分類中,表明利用結構化先驗知識可以提高圖像分類的性能。
- Something-Else: Compositional Action Recognition With Spatial-Temporal Interaction Networks 2020
用於識別訓練集沒有的,對操作物體進行替換的動作
- I Know the Relationships: Zero-Shot Action Recognition via Two-Stream Graph Convolutional Networks and Knowledge Graphs 2019
Action Genome: Spatio-temporal Scene Graphs 2020
- Rethinking Knowledge Graph Propagation for Zero-Shot Learning 2019
- Videos as space-time region graphs.
- Learning to Compare: Relation Network for Few-Shot Learning 2018