jpthu17 / emcl Goto Github PK

作者您好！最近看了您发表的论文，受益匪浅~作为同龄人真的很钦佩！

不知道您在EMCL中有没有试过在MSVD数据集上加DSL（inverted softmax），我尝试在这个数据集上加，但是发现DSL给的源代码是针对方阵的，也就是说文本和视频得是一对一的关系，而MSVD数据集在测试阶段是一个视频对应多个文本。不知道您有没有遇到过这个问题~~

Hello author, thank you for your work. I encountered a problem in the image during the reproduction process. Have you ever encountered it? How was it resolved?

作者你好，我下载了您的代码，尝试去复现论文的结果，但是，我在看代码的时候，发现代码中用的数据是包含audio，face，flow，ocr等特征，但是论文中并没有提到用到这些额外的信息？能帮忙指导一下么？

Thanks for this great work and data sharing~
And when I want to download video data from provided Peking University Yun links, it seems like they are unavailable now. May I ask if these links can be updated to provide downloads? Thanks again!