Comments (4)
No it doesn't. You need to follow the instructions for RealEstate10K from THEIR website. We can't distribute the original frames because (1) there's a huge number -- the memory would be too much -- and (2) for IP reasons.
We then have one change (which we discuss in the file you quoted from above). Their dataset is based on youtube data, which means videos/frames may be missing as hte dataset may change over time. As a result, we save the information about the frames that were correctly extracted.
from synsin.
@oawiles I guess you are not allowed to distribute the extracted frames but I guess you can share the script which you have used to preprocess the data (downloading the video and possibly resize, change intrinsic parameters). I myself have made one but I guess it would be helpful not only for me for comparison but also easier for other people to work faster on your method.
However, I believe that this is just a simple python script that anyone with basic Python skill can do so may be it is not that worth considering anymore :D
from synsin.
Thank you @phongnhhn92 !
We would have given the script but can't distribute that -- it is too similar to simply distributing the frames. But as you say, it is a simple python script that simply extracts frames at the timestamps and resizes them.
from synsin.
What are the png frames to be named individually?
${REAL_ESTATE_10K}/frames/train/${path_totrain_vid1}/*.png
Are they named after their timestamp number in microseconds?
"we create a new text file ${REAL_ESTATE_10K}/frames/train/${path_totrain_vid1}.txt which stores the metadata for each frame (this is necessary as there may be some errors in the extraction process). The ${REAL_ESTATE_10K}/frames/train/${path_totrain_vid1}.txt file is in the same structure as the original text file, except all rows containing images that were not extracted, have been removed."
Explain this step. Does the new txt file have the same 2-6. camera intrinsics and 7-19. camera pose rows as the old txt file?
What is this step for? I would imagine the clip is there to be downloaded and have all its frame extracted or not. Under what circumstance would frames be different?
from synsin.
Related Issues (20)
- increase output image resolution HOT 1
- How to produce the continuously changing video. HOT 1
- module 'pytorch3d.renderer.compositing' has no attribute 'CompositeParams' HOT 5
- Functionality of learn_feature HOT 2
- Neural point cloud renderer HOT 2
- Training on new dataset of rectangular images (H!=W) HOT 4
- Trying to train SynSin on SceneNet database HOT 11
- Discrepancy in sequence of operations inside ResNet block HOT 1
- Inference result does not match with the reported result HOT 3
- Would you recommend me a way to get 360 degree view? HOT 1
- No Downloading Permission of a video HOT 1
- calculation of the relative pose when processing kitti dataset HOT 1
- MIssing project files HOT 1
- Evaluation with a scale factor on RealEstate10K HOT 1
- Some issues to clarify HOT 4
- Is the attribution "lr" controlling learning rate?
- Try to use the code of Appearance Flow in synsin HOT 1
- Wrong setting with depth in PtsManipulator
- Scene episode generation about Matterport3D HOT 13
- habitat-sim install problem
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from synsin.