Evaluation Set

[fuqichen1998]

Hi,

I found the evaluation.py actually use the training images. Could you please share how to get the exact number in Table 3 of your paper. More specifically, how to know which are

the remaining 11 held-out images per time instance for evaluation

?

[zhengqili]

In line 295:

Neural-Scene-Flow-Fields/nsff_exp/evaluation.py

Line 295 in d400175

if img_i % 12 == camera_i:

We skipped the training image by checking if image index % 12 is same as camera index, since there are 12 cameras in total and we generate training images in circular fashion

[fuqichen1998]

Got it, thanks very much for your quick reply!