Skip to content

Latest commit

 

History

History
31 lines (29 loc) · 964 Bytes

demo.md

File metadata and controls

31 lines (29 loc) · 964 Bytes

Demo

Run Sparse Plane Reconstruction on a pair of input images. Download our pretrained model model_ICCV.pth and save to sparsePlane/models

cd sparsePlane/models
wget https://fouheylab.eecs.umich.edu/~jinlinyi/2021/sparsePlanesICCV21/models/model_ICCV.pth

Inference on teaser images.

cd sparsePlane
python tools/inference_sparse_plane.py \
--config-file ./tools/demo/config.yaml \
--input ./tools/demo/teaser \
--output ./debug
  • Predicted correpondence saved as corr.png: teaser corr
  • Reconstruction saved as obj file: obj

Or you can inference on a list of image pairs by making a txt file

cd sparsePlane
python tools/inference_sparse_plane.py \
--config-file ./tools/demo/config.yaml \
--input ./tools/demo/images \
--img-list ./tools/demo/images/image_list.txt \
--output ./debug