-
I saw that the training guide states "Crop videos to 512 * 512 and 25 FPS".
Should I directly use the ffmpeg command you provided to scale the 1080 video to 512 pixels,
or should I recognize a…
-
Thanks for your great work!
When seeing the video, it has a picture of visualization reconstruction scene as following.
![image](https://github.com/user-attachments/assets/b2958b4b-c99e-42e5-8ce3-a…
-
Hi, glad to see such good work!I still have some questions:
Noticed that:
`We sample 220k training samples from our internal video dataset to train our model. ...During training, we randomly sam…
-
Minecraft Version: 1.19.2
Forge Version: 1.253.0-17952
Mod Version: 1.19.2-1.8.4.jar
Single Player or Server: single player
Describe problem (what you were doing; what happened; what shoul…
-
In their example you can load input video, and just drive the mouth and eyes in the video, so it works with full video not just one image
https://liveportrait.github.io/
https://liveportrait.github…
-
I found error load image in replicating LSDA. Error from No faces in frame 300 / MNT/sda1 / video/FaceForensics++ original_sequences/youtube/c23 / videos / 281. Mp4. I re-used the data processing code…
-
**Setup (please complete the following information):**
- OS: Windows 11 22H2
- Player: mpv
- Player Version: 0.35.0-210-ge439ddc0
- Video Quality: 1080p HEVC Dolby Vision Profile 5
**Descri…
-
Why is the DUT output video cropped,Is it because of the model?How should this problem be solved?
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Bug description
I label videos and create a training dataset. however, when I click "train network" the …
-
It have a very hard time when driving video is a bit close and no space above forehead.
I thought it tracked main face features but it might be tracking full head in frame and loses accuracy and gets…