96 citations · 261 across the 9 of their papers we have counts for
Showing 2014Show all
2 papers · 1 filter
cs.LG2014★ 8 cited
Modeling sequential data using higher-order relational features and predictive training
Vincent Michalski, Roland Memisevic, Kishore Konda
Bi-linear feature learning models, like the gated autoencoder, were proposed as a way to model relationships between frames in a video. By minimizing reconstruction error of one fr…
stat.ML2014★ 37 cited
Zero-bias autoencoders and the benefits of co-adapting features
Kishore Konda, Roland Memisevic, David Krueger
Regularized training of an autoencoder typically results in hidden unit biases that take on large negative values. We show that negative biases are a natural result of using a hidd…