基于視覺推理的視頻理解方案

1、基于視覺推理的視頻理解方案技術(shù)創(chuàng)新 變革未來視頻理解典型應(yīng)用機(jī)器人城市管理元人駕駛安防監(jiān)控視頻理解研究目標(biāo)視頻描述的場景？視頻里面的物體？視頻中物體位置？物體間什么關(guān)系？物體與場景關(guān)系？視頻會發(fā)生什么？視頻理解研究代表性任務(wù)Video aptionVideo Question AnsweringMoment RetrievalVideo Relation ReasoningVideo GenerationVideo Grounding視頻理解的挑戰(zhàn)：復(fù)雜場景Kenneth Marino, Ruslan Salakhutdinov, Abhinav Gupta: The More You Kno

2、w: Using Knowledge Graphs for Image Classification. CVPR 2017: 20-28視頻理解的挑戰(zhàn)：小樣本Junyu Gao, Tianzhu Zhang, Changsheng Xu: I Know the Relationships: Zero-Shot Action Recognition via Two-Stream Graph Convolutional Networks and Knowledge Graphs. AAAI 2019: 8303-8311視頻理解的挑戰(zhàn)：可解釋性Will Norcliffe-Brown, Stath

3、is Vafeias, Sarah Parisot: Learning Conditioned Graph Structures for Interpretable Visual Question Answering. NeurIPS 2018: 8344-8353視頻理解的挑戰(zhàn)：Medhini Narasimhan, Svetlana Lazebnik, Alexander G. Schwing: Out of the Box: Reasoning with Graph Convolution Nets for Factual Visual Question Answering. NeurI

4、PS 2018: 2659-2670深度推理學(xué)習(xí)圖神經(jīng)網(wǎng)絡(luò)基本模型Zonghan Wu, Shirui Pan, Fengwen Chen, Guodong Long, Chengqi Zhang, Philip S. Yu: A Comprehensive Survey on Graph Neural Networks. CoRR abs/1901.00596 (2019)ConvGNN for ClassificationGAE for EmbeddingSTGNN for Spatio-temporal LearningConvGNN with Gconvs圖神經(jīng)網(wǎng)絡(luò)應(yīng)用Jie Zhou

5、, Ganqu Cui, Zhengyan Zhang, Cheng Yang, Zhiyuan Liu, Maosong Sun: Graph Neural Networks: A Review of Methods and Applications.CoRR abs/1812.08434 (2018)深度圖推理學(xué)習(xí)：基本要素(a)圖類型(c)訓(xùn)練方(b)傳)方(深度圖推理學(xué)習(xí)到視覺內(nèi)容理解Xinlei Chen, Li-Jia Li, Li Fei-Fei, Abhinav Gupta: Iterative Visual Reasoning Beyond Convolutions. CVP

6、R 2018: 7239-7248復(fù)雜視覺場景Learning+Reasoning基于圖推理學(xué)習(xí)視覺內(nèi)容描述Ting Yao, Yingwei Pan, Yehao Li, Tao Mei: Exploring Visual Relationship for Image Captioning. ECCV (14) 2018: 711-727基于圖推理學(xué)習(xí)視覺內(nèi)容描述Ting Yao, Yingwei Pan, Yehao Li, Tao Mei: Exploring Visual Relationship for Image Captioning. ECCV (14) 2018: 711-72

7、7基于深度推理學(xué)習(xí)的視覺問答Medhini Narasimhan, Svetlana Lazebnik, Alexander G. Schwing: Out of the Box: Reasoning with Graph Convolution Nets for Factual VisualQuestion Answering. NeurIPS 2018: 2659-2670基于深度推理學(xué)習(xí)的視覺問答Medhini Narasimhan, Svetlana Lazebnik, Alexander G. Schwing: Out of the Box: Reasoning with Graph

8、 Convolution Nets for Factual VisualQuestion Answering. NeurIPS 2018: 2659-2670基于深度推理學(xué)習(xí)的視頻動作識別基于卷積的關(guān)系推理Yunpeng Chen, Marcus Rohrbach, Zhicheng Yan, Shuicheng Yan, Jiashi Feng, Yannis Kalantidis: Graph-Based Global ReasoningNetworks, CVPR 2019基于深度推理學(xué)習(xí)的視頻動作識別圖推理單元表示Yunpeng Chen, Marcus Rohrbach, Zhich

9、eng Yan, Shuicheng Yan, Jiashi Feng, Yannis Kalantidis: Graph-Based Global ReasoningNetworks, CVPR 2019圖推理單元可視化結(jié)果基于深度推理學(xué)習(xí)的視覺關(guān)系識別Zhouxia Wang, Tianshui Chen, Jimmy S. J. Ren, Weihao Yu, Hui Cheng, Liang Lin: Deep Reasoning with Knowledge Graph for SocialRelationship Understanding. IJCAI 2018: 1021-10

10、28基于深度推理學(xué)習(xí)的視覺關(guān)系識別Zhouxia Wang, Tianshui Chen, Jimmy S. J. Ren, Weihao Yu, Hui Cheng, Liang Lin: Deep Reasoning with Knowledge Graph for SocialRelationship Understanding. IJCAI 2018: 1021-1028基于深度推理學(xué)習(xí)的組行為識別Jianchao Wu, Limin Wang, Li Wang, Jie Guo, Gangshan Wu: Learning Actor Relation Graphs for Group Activity Recognitio