gengshan-y / expansion

Upgrading Optical Flow to 3D Scene Flow through Optical Expansion, CVPR 2020 (Oral).
https://gengshan-y.github.io/expansion/
MIT License
172 stars 27 forks source link

Running on custom video sequences #5

Open aakash26 opened 4 years ago

aakash26 commented 4 years ago

Hi , Thanks for sharing the code. Is it possible to run on custom video sequences for inference.

Regards Aakash Rajpal

gengshan-y commented 4 years ago

Hi, it's possible. You could place the video frames in the ./input/folder-name/ folder and run

CUDA_VISIBLE_DEVICES=0 python submission.py --dataset seq  --datapath ./input/folder   --outdir ./weights/$modelname/ --loadmodel ./weights/$modelname/$modelname.pth  --testres 1 --fac 2 --maxdisp 512