Open gentleman-zhong opened 2 months ago
May I ask how to use three cameras for training?Thanks
May I ask how to use three cameras for training?Thanks
There are parameters in the configuration file to choose the camera you want to use
May I ask how to use three cameras for training?Thanks
There are parameters in the configuration file to choose the camera you want to use
Is this the sgn_config.py file? It didn't work for me when adding that, how did you do it?
You can directly change the configuration in train.sh @mzlchou
You can directly change the configuration in train.sh @mzlchou
Ok I got it to work sometimes, did you ever get this error?: This is when using cameras 3 4 and 5. 1 and 2 seem to work fine.
sgn_datamanager.py", line 118, in init assert len(self.train_unseen_cameras) > 0, "No data found in dataset" AssertionError: No data found in dataset
You can directly change the configuration in train.sh @mzlchou
Ok I got it to work sometimes, did you ever get this error?: This is when using cameras 3 4 and 5. 1 and 2 seem to work fine.
sgn_datamanager.py", line 118, in init assert len(self.train_unseen_cameras) > 0, "No data found in dataset" AssertionError: No data found in dataset
I encountered the same error, but I used camera 1, which is the default configuration. Did you solve this problem?
Thank you for your open source work! I can render well with one camera, but when I use three cameras, the problem shown in the image occurs. And the problem does not only occur in this scene, but also in other scenes. After a while, the problem of the first picture disappeared and the rendering result appeared normal. What is the reason?