Comments (1)
Well, that's a million dollar question 😉
Overfitting was a big problem in DFDC, and the fact that the training dataset was composed by a few actors surely didn't help.
Adding external data to the training set would have helped but we decided to keep only the original dataset to avoid the risk of being disqualified.
A way to reduce overfitting for the challenge was to split the original dataset into non overlapping subset and train several instances of the same model over those separated sets. You can find a little more in our Kaggle post.
from icpr2020dfdc.
Related Issues (20)
- Facing error while running train_all.sh HOT 2
- Celebdf dataset HOT 1
- Can't find the file data_frame_df.pkl HOT 3
- network doesn't generalize well HOT 3
- Query HOT 4
- Unable to find "List_of_testing_videos.txt" HOT 1
- Source code of Ensemble models HOT 18
- index_dfdc errror HOT 2
- DFDC split HOT 1
- Question about the returned RuntimeError when using the Celeb-DF dataset HOT 7
- what is embedding_routine function used for? HOT 1
- how to set validation split for dfdc? HOT 4
- use pre-trained model to detect my videos, but filter out some specific datasets on the pre-train which may overlap HOT 2
- regarding attention heatmap HOT 1
- Training dataset for DFDC HOT 1
- models pretrained on DFDC works well, but why models pertrained on FFPP always give a low score? HOT 1
- EnvironmentNameNotFound: Could not find conda environment: icpr2020 HOT 3
- About Image prediction notebook on Colab HOT 3
- val_sample HOT 1
- Regarding Top-up Training to improve the accuracy of existing deepfake model
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 icpr2020dfdc.