Open MELANCHOLY828 opened 3 days ago
I think you can just copy the render.py and metric.py from original gaussian splatting, and just simply modify the render_set function to get more output visualization such as depth and normal.
Another thing you need to be aware of is that you need to add else:
behind L81 to ensure the params of trained gaussian point clouds be loaded correctly. (This might be a minor mistake by the author here.)
Can you provide the code for rendering images and evaluation?