Wenyuan Zhang

We noticed that instant-ngp is so fast that the time consumption of quadtree subdivision becomes a significant overhead. Therefore, we adopted error map as a simple subdivision and sampling strategy...

Sorry, I haven't tested it on more advanced works right now. If you have better results on the more advanced frameworks, please feel free to further discuss.

@robofar The explanation from @Xzzit is right. `radii>0` means that this Gaussian is inside the viewpoint, otherwise is outside the viewpoint (the corresponding rgb image).

You can refer to this issue :) https://github.com/griegler/pyfusion/issues/1

> Hi, I'm having the same issue as you mentioned above, did you find a way to solve it? I changed a machine with cuda 11.4 and it worked. But...

@MulinYu Hi, I'm currently running a new experiment with larger training iterations. I will post the result as soon as the experiment completes ;)

This time I trained scannet scene0050_00 for 500k iters. It seems better than that of 30k iters, but is still bad :( I plan to rerun the script using a...

@chenhsuanlin I transformed the data type manually from raw scannet datset. I rescaled the scene so that the center of the scene lies in the coordinate origin (0,0,0) and the...

I did not get a satisfied result at last😥. I'm afraid that maybe there is some generalization ability limit of neuralangelo in indoor scene datatset. If anyone can obtain better...