A deep learning application for gap filling motion capture data.
MoGap looks to use denoising autoencoders to gap fill motion capture data. We use the CMU mo-cap dataset (here: http://mocap.cs.cmu.edu/) and a number of different auto encoder architectures to fill in simulated missing data.
This is still a work in progress but current best results come from our CNN LSTM model which beats state-of-art models run through the same training process. More work needs to be done to validate these results, however.