dvlab-research / SDSD

Seeing Dynamic Scene in the Dark: High-Quality Video Dataset with Mechatronic Alignment (ICCV2021)
52 stars 11 forks source link

The video frame processing code #2

Closed lingyzhu0101 closed 2 years ago

lingyzhu0101 commented 2 years ago

As mentioned in your github description, the .npy file has a resolution of 512 x 960 for fast training. However, the resolution of .png file is 1080 x 1920. The resolutions between them are different.

Could you send the code about resizing and converting .png to .npy? I will be very grateful to your efforts.

My Email : lingyzhu-c@my.cityu.edu.hk