Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Colab end-to-end runnable version #56

Merged
merged 36 commits into from
Apr 18, 2020
Merged

Conversation

AlphaGit
Copy link
Contributor

@AlphaGit AlphaGit commented Mar 23, 2020

Hi!

Along with the help of @styler00dollar, we've created a Python Notebook that can be run in Google Colaboratory for free. We also integrated it with Google Drive and ffmpeg, which means that we can do quick end-to-end tests with input and output files that we carry on.

We're basing our tests from the best model and the DAIN_slowmotion network, but with this, making modifications it's very easy.

We believe that this is a good addition to the original repository, both as a demonstration of how to use the existing code (in real world scenarios) and also as a way for experimenters to run it without the need of a local setup.

We welcome any feedback!

styler00dollar and others added 30 commits March 4, 2020 05:24
- Showing time left
- Showing image progression
- Simplified and remove unused variables
- Simplified cell usage
- Added configuration section at top
Improvements to Colab notebook (Better code, but broken interpolation because of the python file.)
- Improved Collab notebook readme cell
- Added configuration for `FRAME_INPUT_DIR`
- Added configuration for `FRAME_OUTPUT_DIR`
- Showing a bit more info about the GPU Card (some drivers might actually not run)
- Fixed git clone directory path
- Passing `start_frame` and `end_frame` as parameters to the python script (to allow for resuming runs)
- Avoid outputs middle-steps
Added parameters, fixed interpolation estimation, cleaner code, updated readme
@digitalcircuit
Copy link

I downloaded Colab_DAIN.ipynb file from this pull request, uploaded to Google Colab, and followed the steps in the file, and it all worked out! Thank you for formatting this project in a way folks like myself with only an Intel integrated GPU and an old AMD Radeon GPU can experiment with it.

I imagine that once merged, the repository URL inside the notebook should be updated to point to this repository again.

The only issue I had was mere oversight on my part: not paying attention to partial frames, e.g. a 12.5 FPS video does not upconvert in an integer fashion to 60 FPS. 63 FPS is closer to ideal.

@baowenbo baowenbo merged commit 05685d8 into baowenbo:master Apr 18, 2020
vlee-harmonicinc pushed a commit to vlee-harmonicinc/DAIN that referenced this pull request May 14, 2020
Colab end-to-end runnable version
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants