Deep Learning
Train longer, generalize better: closing the generalization gap in large batch training of neural networks
Elad Hoffer, Itay Hubara, Daniel Soudry
We examine the initial high learning rate training phase.
Runtime Neural Pruning
Ji Lin, Yongming Rao, Jiwen Lu, Jie Zhou
The agent judges the importance of each convolutional kernel and conducts channel-wise pruning conditioned on different samples, where the network is pruned more when the image is easier for the task.
Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks
Urs Kรถster, Tristan Webb, Xin Wang, Marcel Nassar, Arjun K. Bansal, William Constable, Oguz Elibol, Scott Gray, Stewart Hall, Luke Hornof, Amir Khosrowshahi, Carey Kloss, Ruby J. Pai, Naveen Rao
Despite advances in limited precision inference in recent years, training of neural networks in low bit-width remains a challenging problem.
Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
Balaji Lakshminarayanan, Alexander Pritzel, Charles Blundell
Quantifying predictive uncertainty in NNs is a challenging and yet unsolved problem.
Imagination-Augmented Agents for Deep Reinforcement Learning
Sรฉbastien Racaniรจre, Theophane Weber, David Reichert, Lars Buesing, Arthur Guez, Danilo Jimenez Rezende, Adriร Puigdomรจnech Badia, Oriol Vinyals, Nicolas Heess, Yujia Li, Razvan Pascanu, Peter Battaglia, Demis Hassabis, David Silver, Daan Wierstra
Neural Information Processing Systems http://nips.cc/
Learning Disentangled Representations with Semi-Supervised Deep Generative Models
Siddharth N, Brooks Paige, Jan-Willem van de Meent, Alban Desmaison, Noah Goodman, Pushmeet Kohli, Frank Wood, Philip Torr
We evaluate our framework's ability
Learning a Multi-View Stereo Machine
Abhishek Kar, Christian Hรคne, Jitendra Malik
We present a learnt system for multi-view stereopsis.
Revisit Fuzzy Neural Network: Demystifying Batch Normalization and ReLU with Generalized Hamming Network
In particular, we conjecture and empirically illustrate that, the celebrated batch normalization (BN) technique actually adapts the "normalized" bias such that it approximates
Z-Forcing: Training Stochastic Recurrent Networks
Anirudh Goyal ALIAS PARTH GOYAL, Alessandro Sordoni, Marc-Alexandre Cรดtรฉ, Nan Rosemary Ke, Yoshua Bengio
Stochastic recurrent models have been successful in capturing the variability observed in natural sequential data such as speech.
Teaching Machines to Describe Images with Natural Language Feedback
Robots will eventually be part of every household. It is thus critical to enable algorithms to learn from and be guided by non-expert users.