arielbaz

5 ms·

1. 1.▲
   by arielbaz 6y ago

   see also "Training a deep learning model for deepfake detection" ( https://news.ycombinator.com/item?id=22433711 )
2. 2.▲
   Training a deep learning model for deepfake detection(blog.claysciences.com)

   10 points by arielbaz 7y ago|0 comments
3. 3.▲
   by arielbaz 8y ago

   Is combosaurus one of the 10,000? https://www.quora.com/Whatever-happened-to-Combosaurus
4. 4.▲
   A comprehensive guide to annotating videos for Machine Learning(linkedin.com)

   9 points by arielbaz 8y ago|0 comments
5. 5.▲
   by arielbaz 8y ago

   First, thanks for sharing this data. Second - why on earth would anyone create a 1.8TB zip file of 100k videos? Likely the video encoder already compressed every possible bit out of these videos, zip is not going to make it smaller . It is
6. 6.▲
   Why annotating data directly on videos is way superior to annotating images(linkedin.com)

   14 points by arielbaz 8y ago|0 comments
7. 7.▲
   Machine Learning: garbage in, garbage out? Train with highest quality data(linkedin.com)

   24 points by arielbaz 8y ago|0 comments