Questions about points fusion on ETH3D dataset

[progressforever]

Hello! Thank you for your source code~

I want to ask some questions about points cloud fusion on ETH3D dataset. Because the data of ETH3D dataset is gray image with channel of 1. However, the DTU dataset is RGB image with channel of 3. How can I generate points cloud on ETH3D dataset using the pretrained model on DTU model?

I use (cv2.imread(gray.png)) to let the channel of gray image 3. But the result is very bad!

Can I transfrom the RGB image to gray image on DTU and get a pretrained model using gray DTU dataset? Will it works?

Hope your early reply! Thank you

[YoYo000]

I directly use the test.py to generate the depth map. And use visibility+average fusion to fuse the depth maps. The Gipuma fusion since not work well on ETH3D scenes.

I think retraining with gray images may work but I haven't tried this before.