JiawangBian / SC-SfMLearner-Release

Unsupervised Scale-consistent Depth Learning from Video (IJCV2021 & NeurIPS 2019)
https://jwbian.net/sc-sfmlearner
GNU General Public License v3.0
733 stars 149 forks source link

about inverse_warp.py #100

Open JGGGBang opened 6 months ago

JGGGBang commented 6 months ago

I would like to inquire about the differences between these two functions: ①you:def inverse_warp2(img, depth, ref_depth, pose, intrinsics, padding_mode='zeros') ②cv2.warpPerspective(src,M, dsize[,dst[,flags[, borderMode[ , borderValue]]]]) →dst

JiawangBian commented 6 months ago

at least, the cv2 functions are not differentiable :)

JGGGBang commented 6 months ago

In fact, what I want to achieve is to find the newly added part of frame i+1 relative to frame i of the image