An implementation of Stereo Magnification: Learning View Synthesis using Multiplane Images in tensorflow
- The original network performs planar transformation on the predicted color images and then alpha-composit them. In current version, The planar transformation part has been ignored.
I organized my code according to that of this project:Adversarial Video Generation, which makes it easy to maintain. Thanks to dyelax.