• Publications
  • Influence
Pruning Convolutional Neural Networks for Resource Efficient Inference
TLDR
We propose a new criterion based on Taylor expansion that approximates the change in the cost function induced by pruning network parameters. Expand
  • 564
  • 78
Compressing Neural Networks with the Hashing Trick
TLDR
We present a novel network architecture, HashedNets, that exploits inherent redundancy in neural networks to achieve drastic reductions in model sizes. Expand
  • 771
  • 45
  • PDF
Online Detection and Classification of Dynamic Hand Gestures with Recurrent 3D Convolutional Neural Networks
TLDR
In this paper, we address these challenges with a recurrent three-dimensional convolutional neural network that performs simultaneous detection and classification of dynamic hand gestures from multi-modal data. Expand
  • 253
  • 42
  • PDF
Pruning Convolutional Neural Networks for Resource Efficient Transfer Learning
TLDR
We propose a new framework for pruning convolutional kernels in neural networks to enable efficient inference, focusing on transfer learning where large and potentially unwieldy pretrained networks are adapted to specialized tasks. Expand
  • 250
  • 31
  • PDF
Learning with Marginalized Corrupted Features
TLDR
We propose to corrupt training examples with noise from known distributions within the exponential family and present a novel learning algorithm, called marginalized corrupted features (MCF), that trains robust predictors by minimizing the expected value of the loss function under the corrupting distribution. Expand
  • 135
  • 21
  • PDF
Reinforcement Learning through Asynchronous Advantage Actor-Critic on a GPU
TLDR
We introduce a hybrid CPU/GPU version of the Asynchronous Advantage Actor-Critic algorithm, currently the state-of-the-art method in reinforcement learning for various gaming tasks. Expand
  • 96
  • 17
  • PDF
Non-linear Metric Learning
TLDR
We introduce two novel metric learning algorithms, Χ2- LMNN and GB-LMNN, which are explicitly designed to be non-linear and easy-to-use. Expand
  • 166
  • 12
  • PDF
Parallel boosted regression trees for web search ranking
TLDR
We propose a novel method for parallelizing the training of GBRT, based on data partitioning, that achieves almost perfect linear speed-up with up to about 48 cores on the large data sets. Expand
  • 129
  • 10
  • PDF
Importance Estimation for Neural Network Pruning
TLDR
We propose a novel method that estimates the contribution of a neuron (filter) to the final loss and iteratively removes those with smaller scores. Expand
  • 93
  • 10
  • PDF
Stochastic Neighbor Compression
TLDR
We present Stochastic Neighbor Compression (SNC), an algorithm to compress a dataset for the purpose of k-nearest neighbor (kNN) classification. Expand
  • 56
  • 9
  • PDF