Open harshilcontineu opened 7 months ago
Hi, Nice to hear from you.
Yes, you are right. Directly using the current solution may not deal with pillars, furniture, or any large occlusion in the middle of the rooms. The main limitation is due to the layout model because the data domain used during training differs from the scene application. You can check this 360-mlc.
You can try the following: NO TOO EASY OPTIONS:
EASY OPTION:
Hey First of all, great implementation! so thank you for that.
I was using this to create floorplans of spaces with thick rectangular pillars in the middle of it. Does this implementation account for such objects? I'm assuming not but I wanted your thoughts on it.
Thanks Harshil