Closed MrinalJain17 closed 6 years ago
Creating a function for saving the video tensor does not seem to be the best option. This could be achieved by the user by using the np.save or np.savez or other similar functions. A function for the same will clutter the library. But with that said, this will be referenced in the documentation in the next release.
Writing videos (altered) back to the disk does make sense. But, there is certainly a lot of uncertainty:
Further, the user would have to again go through the process of reading videos in order to use them again, which could be time-consuming. Therefore, although this functionality could be useful, it will not be implemented in favor of the first method.
Support to save the video tensor. Two different functionalities could be provided for this feature: