EPFL-VILAB/omnidata
A Scalable Pipeline for Making Steerable Multi-Task Mid-Level Vision Datasets from 3D Scans [ICCV 2021]
Jupyter NotebookNOASSERTION
Issues
- 1
Train own dataset
#62 opened by rick-dn - 1
Download the dataset failed
#59 opened by carolhwchen - 19
Camera intrinsics
#29 opened by VitorGuizilini-TRI - 6
normal v2 pretrained model not working
#53 opened by yuyingyeh - 3
Can't download blendedmvg data
#48 opened by TeCai - 1
Fail to download subset 'tiny'
#61 opened by Hermera - 0
Google Objects dataset
#60 opened by miangoleh - 1
Potential bug in depth estimation training code
#41 opened by JasonQSY - 3
Failed to retrieve pretrained models
#52 opened by dengnianchen - 2
- 4
- 7
- 0
- 0
Checkpoint for DPT-Large for depth estimation
#57 opened by adondera - 2
Doubt about the cosine_angular_loss
#47 opened by ireneMsm2020 - 0
Scene metadata Google drive link not working
#55 opened by woodfrog - 2
Inverse in Midas Loss
#50 opened by jamekuma - 3
- 0
Pitch and Roll values?
#54 opened by miangoleh - 3
Surface normal model
#23 opened by hit421 - 0
Depth Estimation with Kitti Images
#51 opened by aartykov - 2
- 4
Download scripts fail with socket.timeout
#44 opened by janbechtold - 2
Can't download semantic data.
#40 opened by HLinChen - 2
Pose consistency in image sequences
#46 opened by VitorGuizilini-TRI - 1
why not include scannet v2 models ?
#45 opened by pengwangucla - 1
- 2
training code of V2
#25 opened by Tord-Zhang - 9
- 0
question about the depth estimation output
#42 opened by Twilight89 - 3
How to clip predictions for evaluation
#34 opened by jeeyung - 1
Depth output
#39 opened by stalkerrush - 0
Depth values and visualization
#38 opened by DianCh - 1
depth2normal issue
#31 opened by hit421 - 1
- 0
- 2
how to download omnidata_dpt_depth_v2.ckpt file?
#22 opened by saiedg - 1
some question about image scale
#21 opened by LiXinghui-666 - 1
- 9
Different results between using the pretrained model locally and using the web demo
#19 opened by ryunuri - 1
download pretrained models
#30 opened by yxt7979 - 0
- 0
Surface normal map output camera direction?
#27 opened by dxu - 1
- 1
What if I put the 512*512 image to this model...
#17 opened by lxl1115 - 1
About the normal coordinate...
#18 opened by lxl1115 - 3
Output normal vector not normalized
#14 opened by stalkerrush - 1
how to change 3 channel depth to 1 channel
#15 opened by zf-666 - 11
- 1