-
ghost updated
1 month ago
-
This is a great work! you said "Then, we randomly select the constructed point cloud of the video frames and render it along the estimated camera trajectory using Pytorch3D." in Sec. 4.1, from my pers…
-
So I was looking over this to see how hard it would be to finish these up. Hats/Headbands tend to be fairly easy, as is anything on the head. As long as I have cutouts we're good.
The biggest issue…
-
Samples: [GitHub](https://github.com/user-attachments/files/17535949/scroll.marquee.performance.tar.gz), [0x0.st](https://0x0.st/XURV.tar.xz) (I uploaded to 0x0.st before deciding to make this a GitHu…
-
Hi, @li-ronghui
When I did python dld/data/pre/FineDance_normalizer.py to get normalizer.pth, I got 822,177 frames from ./data/finedance/mofea319(203 *.npy files), not available 831600 frames mentio…
-
Hi, first of all, many thanks for this great tool, really helped me to visualize what's going on in our app 💪🏻
However I wanted to ask how to use ref frames? I've found just short mention in the docs…
-
Hello,
I use the Noetic ROS driver for an SBG Ellipse D. I noticed several inconsistencies in the /tf management.
First of all, we are asked to name 4 frames in the config file, so I expected to…
-
Hello! First, I want to express my gratitude for HBC, MAS, and FC.
I have a feature request for MAS that I believe is straightforward. Would it be possible to add an option to render selected actio…
-
您好！
我想问一下关于RGB image的问题，您是直接将得到的person frame拼接成一张图片吗？就是每一个视频最后就只得到一张图片是吗？
-
See existing work in #812, and see below for a list of functions that we needed to consider, and some thoughts on what form of genericity is needed. Goal is to make sure that data frame extensions ret…