Renmin University of China and Zhejiang Gongshang University at TRECVID 2019: Learn to Search and Describe Videos
Renmin University of China and Zhejiang Gongshang University at TRECVID 2019: Learn to Search and Describe Videos
复制标题
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Xirong Li;Jinde Ye;Chaoxi Xu;Shanjinwen Yun;Leimin Zhang;Xun Wang;Rui Qian;Jianfeng Dong
中科院分区：
文献类型：
--
作者：
Xirong Li;Jinde Ye;Chaoxi Xu;Shanjinwen Yun;Leimin Zhang;Xun Wang;Rui Qian;Jianfeng Dong
在本文中，我们总结了我们的TRECVID 2019 [1]视频检索实验。我们参与了两个任务：Ad-hoc视频搜索（AVS）和视频到文本（VTT）。对于AVS任务，我们基于两种深度学习模型开发解决方案，即：W2 VV ++网络[11]和双编码网络[7]。对于VTT匹配和排名子任务，我们的条目也基于W2 VV ++和双重编码网络。对于VTT描述生成子任务，我们通过多级视频编码和属性预测来增强经典的编码器-解码器模型。2019年版TRECVID基准测试对我们的联合团队来说是一次富有成效的参与。根据ciderD标准，我们的运行在AVS和VTT匹配和排名任务中排名第二，在VTT描述生成子任务中排名第三。
In this paper we summarize our TRECVID 2019 [1] video retrieval experiments. We participated in two tasks: Ad-hoc Video Search (AVS) and Video-to-Text (VTT). For the AVS task, we develop our solutions based on two deep learning models, i.e . the W2VV++ network [11] and the Dual Encoding Network [7]. For the VTT Matching and Ranking subtask, our entry is also based on the W2VV++ and Dual Encoding Networks. For the VTT Description Generation subtask, we enhance the classical encoder-decoder model with multi-level video encoding and attribute prediction. The 2019 edition of the TRECVID benchmark has been a fruitful participation for our joint-team. Our runs are ranked at the second place for AVS and VTT Matching and Ranking tasks and the third place for the VTT Description Generation subtask in terms of the ciderD criterion.