JialinKang

Johns Hopkins University the U.S.

Results 6 issues of


                                            JialinKang

Add a video inpainting paper

great conclusion! hope to update immediately there is a new paper for video inpainting `Decoupled Spatial-Temporal Transformer for Video Inpainting`

How to generate the groundtruth optical flow by myself in endoscopy image?

4

comment

How to solve this problem `THC.h:No such file or directory` when run the command 'bash install_scripts.sh'?

2

comment

![image](https://user-images.githubusercontent.com/54913794/108458726-2bd32300-72b0-11eb-81c6-549270e24198.png)

Is this deep image blending algorithm faster than Possion blending?

is the frame the ground truth? but the paper says there is no groundtruth for training

2

comment

https://github.com/ruiliu-ai/DSTT/blob/0b16ff17afbfabda73517903524e50eccc79defe/core/trainer.py#L252

Could you write a tutorial to introduce the usage of this repository?