Skip to content
Dominik edited this page Jul 16, 2015 · 14 revisions
Task
Reading papers about state of the art, introductions to deep learning all
Set-up caffe on local machines, investigate research branches all
Image cleansing Joseph, Tanja
Image prediction with c++ and python Tanja, Stefan, Dominik, Tom B.
Set-up RNN Stefan, Dominik, Tom B., Tanja
Extracting optical flow images Tom H., Joseph, Tom B.
Train spatial CNN Tanja, Joseph
Script to transform data into lmdb Tom H., Tom B.
Script to import stacked flow images into lmdb Tom H., Joseph
Introductory presentation all
Masterproject presentation all
Fine tune existing CNN to UCF-101 Dominik, Stefan, Tom B.
Train flow CNN Tanja, Joseph, Tom H., Tom B.
Set up new server all
Create web demo Tom H., Tom B.
Build HDF5 DB Tom B.
Build LevelDB Dominik, Stefan, Tanja
Extracted cropped frames from videos Joseph, Tanja
Create LSTM on top of CNN_M and LMDB Dominik, Stefan
Created simple fusion network Tanja
Get the Flow network to train above 5 % Tanja, Dominik, Stefan
Train fusion network Tanja, Stefan, Dominik
SVM fusion Joseph, Dominik
Clone this wiki locally