sparse-to-dense
sparse-to-dense copied to clipboard
ICRA 2018 "Sparse-to-Dense: Depth Prediction from Sparse Depth Samples and a Single Image" (Torch Implementation)
I am trying to apply your model to slam application. I am using a stereo setup with wide angle camera lenses to obtain the sparse depth samples. With the pretrained...
Hi @fangchangma Hope you are doing well during these tough days. I am a student studying the Netherlands currently, working on a project that involves NYUv2 RGBD dataset. My question...
Hi @fangchangma , In your code, you regard the number of all pixels as "utils.nValidElement" when evaluation AbsRel and others, rather than the number of Lidar-point. Actually, I want to...
(base) by1706143@BY1706143:/tmp/mozilla_by17061430/sparse-to-dense-master$ th main.lua -testOnly true -dataset nyudepthv2 -inputType rgbd -nSample 0 -rep linear -encoderType conv -decoderType deconv2 -criterion l1 -LR 0.01 -pretrain true error: trained model does not exist:...
error: trained model does not exist: /tmp/mozilla_by17061430/sparse-to-dense-master/results/nyudepthv2.input=rgbd.nsample=100.rep=linear.encoder=conv.decoder=upproj.criterion=l1.lr=0.01.bs=16.pretrained=true/model_best.t7 Usage: [options] Sparse-to-Dense Training script See https://github.com/fangchangma/sparse-to-dense for details Train Options: -dataset Options: nyudepthv2 | kitti [nyudepthv2] -inputType Options: rgb | rgbd |...
We want to kown,how do you center-cropped image?Thank you.
Based on NYU v2,the original frames of size 640×480 are first down- sampled to half and then center-cropped, producing a final size of 304×228. we have a question.How is this...
Thanks for your perfect project. The question is when I use following script to read ht5 file, the rbg image I get is a whole white image. How to recover...
Thks for sharing your perfect job. I'm interested in your dense-map from VOSLAM. But I don't know how to generate sparse depth image. Can you provide your processed dataset generated...
Hi @fangchangma, Thanks for the great work. Is it possible to upload the dataset to Google Drive for downloading? Since the download speed is limited to only a hundred KB/s...