is there any implementation of 3D partial convolution in keras or tf

[wooramkang]

now i'm trying to make some codes to build a model for video inpainting, not image inpainting. on researching inpainting models, i found partial convolution works well so i would like to build a model with partial convolution for 3D but there is no standard codes or codes working well that i can follow

so do your team have a plan to do or doing some works or research for it? so is there any reference to read some codes, not papers?

what i mentioned) https://github.com/MathiasGruber/PConv-Keras https://arxiv.org/pdf/1804.07723.pdf

[liuguilin1225]

We haven't worked on 3D yet. You can also find some other video inpainting code repos. There are some recent works on video inpainting in CVPR/arxiv.