leonhlj / fac-net Goto Github PK

Hello, Thanks for making your interesting work open to the public,
I am a student recently studying on WS-TAL tasks.
What I want to ask is about "how to visualize results from network so that it can be the one shown in the paper."
For example, like the ones in the Figure 1 and 5 in your paper.

Thanks, Joseph

你好，我是一个学生，最近在复现您的论文，但是我使用您的代码，训练了50000轮后，mAP在任何iou下都降为了0，请问您遇到这样的情况了吗

您好！
我在您的utils/video_dataloader.py文件里面，有看到-I3D-JOINTFeatures.npy文件：

或者我如何制作一个总的ActivityNet1.3-I3D-JOINTFeatures.np文件以及ActivityNet1.3-Annotation文件夹里面的classlist.npy、labels_all.npy等文件呢？

冒昧打扰，十分抱歉！

请问这篇文章是ICCV21的文章吗？ @LeonHLJ

The first three lines are the result of my own training model：
100 100 34.18 28.70 20.20 13.54 8.17 3.40 1.45
35 70 31.39 26.72 19.16 12.83 7.28 3.66 1.70
35 75 31.38 26.49 19.76 13.25 7.76 3.84 1.81
yours:
1000 1000 68.19 62.43 53.21 44.78 34.10 23.13 13.44

What is the moment localization network used to achieve the accuracy in this paper

Hello, Thanks for making your interesting work open to the public.
Could you please provide a requirements.txt?

Thanks for your job, Could you provide the Baidu cloud link of extracted features